Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2006-08-08

00:00:31lamedi ment the list of course
00:00:57Paul_The_Nerdsenab: Yes
00:01:01*amiconn doesn't follow the lists too closely nowadays
00:01:16lamedit was about a week ago
00:01:30senabPaul: aaaah ok, now i know ;)
00:02:09amiconnHmm, +386 bytes on archos recorder. less than I thought, but I still consider it a suboptimal solution
00:02:38 Quit cq ("Chatzilla 0.9.75 [Firefox 1.5.0.6/2006072814]")
00:02:47 Quit lightyear (Remote closed the connection)
00:05:30 Quit senab (" HydraIRC -> http://www.hydrairc.com <- The future of IRC")
00:06:51lamedamiconn: I still think it's a good solution until all screens would use the user font, and that might even keep us informed in what there is to do both at the same time having users using their own native language. furthermore I think I'll start by trying to port the pitchscreen, do you have refrences of the way that you think it should be done? 2.do you vaguely remember something about holding left/right for horizontal scrolling and do you have an
00:08:59amiconnI think on the pitch screen you'll run into the same problem as with the quickscreen - you cannot place 2 texts next to each other and have them scrolling within their box
00:09:12amiconnThis will need viewports.
00:09:29amiconnI guess the recording screen will be easier, evenn though it contains more
00:09:58peturamiconn: gray_core.c:1473: warning: unused variable 'mask' (see cvs build table)
00:10:07amiconnyes, saw that
00:11:09amiconnFor some reason I overlooked it in my local tests
00:11:18lamedthe pitch screen is using the GUI_QUICKSCREEN widget.. and I have the smallest idea about what you are all talking about with the viewports so I couldn't start working on something like that.
00:11:21*linuxstb commits
00:12:27bluebrotherhow about truncating long strings for now? We could just replace the last characters with "..."
00:12:47lamedanyone wants to express his thoughts about http://www.rockbox.org/tracker/task/5182?
00:13:24amiconnlamed: The problem I am talking about is that both the quickscreens and the pitch screen have actions associated with << and >>, and the texts are placed resembling the button "cross" / joystick
00:13:33lamedbluebrother: are you refering to my issue?
00:13:39bluebrotherlamed, yes.
00:14:31amiconnSo there are 2 texts to be placed next to each other, i.e. into th eleft & right half of the screen. Now each of the texts can be longer than 1/2 screen width and would need to scroll - within its area
00:14:51amiconnThat's currently impossible, but will be, hopefully soon
00:15:22bluebrotherI was previously thinking about something like that for a complete different topic. When not using scrolling lines in wps the line gets simply truncated so having them clipped with a trailing "..." like e.g. konqueror does would be nice.
00:15:50lamedamiconn: I guess that would take changing the scrolling thread?
00:15:55amiconnbluebrother: How would you fit "..." onto a string that can only be 3 chars long?
00:16:20Paul_The_NerdIsn't there defined a ... character?
00:16:25bluebrothercould be problematic.
00:16:37lamedbluebrother: the problem in my case is non-latin language users get squares instead of text where rockbox is using the internal font
00:16:56amiconnlamed: Not only the scroll thread. The idea is to introduce viewports, i.e. selectable screen areas which behave like a window for drawing operations, with clipping and all
00:17:03bluebrotherbut usually that problem would occur with bigger sized displays and big fonts I guess.
00:17:33bluebrotherlamed, thats true but I was thinking about a way to get the quickscreen using the user font without scrolling
00:18:03amiconnThe scrolling in the quickscreen is broken anyway, because of another deficiency in the scroll thread
00:18:05bluebrotherhaving viewports would be better, but maybe there's a simple not-so-ideal solution for now.
00:18:38amiconnbluebrother: I have a problem with half-baked solutions. They tend to block the proper solutions
00:18:39lamedamiconn: well then, I guess i'm back at where I started, thinking it's better having things my way up until one of you gurus would have viewports up and running :P
00:19:15*amiconn doesn't consider viewports higher magic
00:19:31amiconnThey will make a number of things easier
00:20:16*amiconn thinks that viewports should be his next project
00:20:18Paul_The_NerdNot to mention they've already been implemented.
00:20:24amiconnPaul_The_Nerd: ??
00:20:38Paul_The_Nerdamiconn: Either Bluechip or that other guy implemented them.
00:20:42lamedjust agree you're one of the project's gurus, that's it :) and about the viewports, please try to remember we are missing RTL scrolling for hebrew (and arabic but I don't think anyone is using arabic with RB)
00:20:44bluebrotheramiconn, I agree with you about half-baked solutions not being good. But there's always a tradeoff between solutions and maybe sometimes a not-optimal solution could be cheaper.
00:20:46Paul_The_NerdI think it was the other one.
00:20:50amiconnAre you referring to the scrolling margins patch? That's not viewports
00:20:56Paul_The_NerdNo, actual viewports supposedly
00:21:00Paul_The_NerdI think it was gl.tter.
00:21:22bluebrotherthus I don't consider it as useless discussing it ;-)
00:21:23Paul_The_NerdHe posted to the mailing list about his personal build that he was gonna be maintaining, and I went and checked it out, and it claimed to have viewports implemented.
00:21:34amiconnIn fact I already have the basics in my head, and given that I wrote most of the drawing code...
00:21:52Paul_The_NerdIt's probably better that we have our own implementation of them anyway. :)
00:22:32*linuxstb breathes a sigh of relief at the green build table
00:22:50amiconnHmm :/
00:23:03lamedbluebrother: I'm going to take a shot editing your nice manual :)
00:23:20amiconnlinuxstb: Why did you #ifdef lcd_yuv_blit() for colour ipods? There are more colour targets...
00:23:45amiconn(in lcd.h)
00:24:06bluebrotherlamed, my manual? I'm not the only guy working on it ... I guess I'm more the latex guy than a text writer
00:24:13linuxstbBecause it's only implemented for those.
00:24:46amiconnDifferent philosophy then...
00:24:47lamedblue: I just always see your name when it gets edited or something :)
00:26:24Paul_The_Nerdamiconn: You'd rather it be defined but empty for the other targets, so that people know to fill it in?
00:27:18bluebrotherlamed, I guess this is mainly because RL sometimes keeps me away from working on it for quite some time, and when that's over I have some stuff to do ;-)
00:27:34amiconnPaul_The_Nerd: I would just have it declared for other targets that _should_ implement it. As long as it's not used, the declaration doesn't hurt
00:27:39lamedwow, editing the beast looks bad.
00:29:42bluebrotherlamed, what beast?
00:33:01lamedwhat's the button hold macro..?
00:34:02bluebrotherhmm ... there isn't any.
00:34:40bluebrotherI guess because hold isn't a "button" in that sense. At least it was up to now.
00:35:36bluebrotherbut I guess a sentence like "when hold is enabled the game is paused" wouldn't be problematic. Mainly because hold is somewhat a special key.
00:37:12amiconnNot all targets have a hold switch. Some places use "soft hold" with an ordinary button (combo) on these targets
00:37:31*amiconn wonders how that could be handled nicely in the manual
00:37:53amiconn..especially as some plugins do use the hold switch as a button on
00:38:20bluebrotherhmm. Wasn't aware of that.
00:38:27amiconn(doom on ipod uses it to bring up the in-game menu, and rockboy on h1x0 uses it to flip through the various scaling options)
00:41:37 Part Paul_The_Nerd
00:43:44 Join razorblade [0] (n=c9f34820@labb.contactor.se)
00:43:57razorbladehello everyone!
00:44:02 Quit EbErT ()
00:44:10lamedhi!
00:44:48lamedbluebrother: you where right on it, I wanted to add a HOLD pause games message.
00:44:48razorbladecan anyone help me? i have a very annoying issue while compiling rockbox :-/
00:45:05lamedwell go ahead and ask, razorblade!
00:45:49razorbladeright, i have an ipod photo, and i just wanted to compilate the source becouse i wanted to try the wheel patch of the rockboy you know
00:46:02lamedbluebrother: what do you want to do about it? make a new macro?
00:46:12bluebrotherlamed, I guess we need to find a good solution for the hold issue in the manual. I've added it to ManualTodo for now.
00:46:17razorbladei used the script in the sources tools, to make everithing i need, i run the configure
00:46:42bluebrotherlamed, I'm not sure. I guess we also need to describe the hold function somewhere shortly.
00:46:55razorbladebut when i "make", it seeems to have a problem while compilating the sysfont.o, it says something about "long calls" not allowed
00:47:05razorbladeand the compilations just stops there
00:47:22linuxstbWhich version of arm-elf-gcc are you using?
00:47:33bluebrotherif you want to add something now just mark it with the \fixme macro. See LatexGuidelines for some explanations.
00:47:50razorblade2.95.3 if i remember well
00:47:55razorbladeand gcc 4.0.3
00:48:16linuxstbThat's why - you need arm-elf-gcc 4.0.x as well.. Or at the very least 3.4.x
00:48:25amiconnrazorblade: gcc 2.95.3 is way too old for anything rockbox
00:48:34razorbladehmmmmm
00:48:36amiconnFor ipod you will need 4.0.x
00:48:48bluebrotherlamed, got to go now. Back another day.
00:49:00lamedbluebrother: nah, I'm fed up with doing unneccecery work for one day :/
00:49:06lamedgood night
00:49:17Bagderrazorblade: use the rockboxdev.sh script and it'll pick the proper version for you
00:49:22lamed(as long as it's on your todo list)
00:49:27 Quit bluebrother ("bye")
00:49:29razorbladeyeah! i did that
00:49:46razorbladeit took long time before downloaded everithing
00:49:58razorbladebut seems like i need arm-elf 4.0.x
00:50:03Bagderthen you have 4.0.3
00:50:03razorbladehmmm is there any binary package there for ubuntu/debian?
00:50:08Bagder?
00:50:16Bagderrockboxdev.sh builds arm-elf-gcc
00:50:18Bagderand friends
00:50:36razorbladei used it
00:50:56linuxstbMaybe you have arm-elf-gcc 2.95.3 in your $PATH as well (before 4.0.x)
00:51:29amiconnThe arm multiplier is a strange thing...
00:52:04amiconnFrom what I read, it's a quite cheap solution, and multiplies byte-wise, making execution time depend on the input data
00:52:23razorbladehmmmm have sense... well the sure thing is, i used the script in "tools" but when running configure says Using arm-elf-gcc 2.95.3 , so i need to find the way to change that
00:52:24amiconn(the cpu optimises multiplication by zero away)
00:53:20Bagderrazorblade: as linuxstb said, edit your path
00:54:52razorbladeok, im doing it alredy lets see what can i do
00:56:18bilbravoBagder: are you still working on the Sansa port?
00:56:29Bagderyes
00:56:35 Join Paul_The_Nerd [0] (n=Paul_The@cpe-70-112-165-230.austin.res.rr.com)
00:56:38Bagderjust slowly
00:56:42bilbravoheh
00:56:54bilbravowell, i'd be interested in helping
00:57:07bilbravohowever, no rockbox experience other than installing it on my fiance's ipod
00:57:10Bagderyou'll be very welcome!
00:57:29bilbravogreat
00:57:40bilbravoI tried to send you an e-mail, but it was returned!
00:57:55 Quit ender` (Read error: 104 (Connection reset by peer))
00:58:06Bagdertry again
00:58:11 Quit mirak (Connection timed out)
00:58:13BagderI get thousands of mail every day
00:58:34bilbravowell, basically it was asking if you were still working on it and offering help, so i don't think that one is necessary :-)
00:58:53Bagderyou seen my latest page I put up today?
00:59:01 Quit mkey (""Welcome to IRC; Where men are men, women are men and little girls are FBI agents!"")
00:59:03Bagderhttp://daniel.haxx.se/sansa/e200-devboard.html
00:59:18bilbravoi have not, i'll look now
00:59:54razorbladesweet! thank you guys, that problem was about to break my integrity hehehe
00:59:58Paul_The_Nerdbilbravo: He often posts mention of updates to the Sansa thread in the forum,.
01:00
01:00:30bilbravoah, haven't perused the forums much... just saw the note about the Sansa port on anythingbutipod.com
01:00:34bilbravothought i'd offer a lending hand
01:00:58bilbravoas i'd love to play ogg/vorbis on my e250, plus why leave something alone wheni can mess with it?
01:00:58Paul_The_NerdIt's early enough along that there's still plenty to do. ;)
01:01:11Bagderoh yes
01:03:42Bagderbilbravo: the main focus right now for us is to figure out how to use the LCD
01:03:51bilbravook
01:03:52Bagderus being me and MrH
01:03:55bilbravoi'm reading through the forum
01:04:15bilbravoi'll be leavingmomentarily, but i'll catch you tomorrow sometime or later tonight (not sure your timezone)
01:04:25Bagder1 am here
01:04:33BagderI'm off really soon
01:07:18***Saving seen data "./dancer.seen"
01:08:26bilbravoah, 7 PM here :-D
01:08:28 Join thegeek [0] (n=thegeek@s026b.studby.ntnu.no)
01:13:33razorbladedamnit, the whole thing is screwed up!
01:13:47 Join mkey [0] (n=mkey@pD9E354E7.dip.t-dialin.net)
01:14:06razorbladeare there any compilated build for an ipod photo, with the wheel fix for the rockboy?
01:14:17razorblade*builds
01:14:20 Quit dan_a (Read error: 110 (Connection timed out))
01:14:40linuxstbThere are one or two "experimental builds" in the forums, but I don't know if they include that patch.
01:16:07razorbladehmmm crap i've alredy tryed those... well seems like im f%&ked up :( thank dudes anyway (Y)
01:18:43razorbladeanyway, my sistem was kinda unstable, i thing im gonna reinstall everything
01:22:39 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
01:26:51 Quit mkey (""Welcome to IRC; Where men are men, women are men and little girls are FBI agents!"")
01:27:48 Quit razorblade ("CGI:IRC")
01:29:48 Quit Arathis (Read error: 110 (Connection timed out))
01:35:11 Part Paul_The_Nerd
01:39:18 Quit mikearthur (Remote closed the connection)
01:40:14 Quit amiconn (Read error: 110 (Connection timed out))
01:40:14 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
02:00
02:00:48 Quit petur ("here today, gone tomorrow")
02:00:55 Quit daurn|afk (Read error: 110 (Connection timed out))
02:02:37 Join spiorf [0] (n=spiorf@host47-143.pool8250.interbusiness.it)
02:03:20 Quit spiorf (Client Quit)
02:09:14 Join Paul_The_Nerd [0] (n=Paul_The@cpe-70-112-165-230.austin.res.rr.com)
02:10:27netmasta10btBagder: have u tried 'manufacturing mode' on the 260r ?
02:15:47 Quit JoeBorn ("Leaving")
02:21:30preglowamiconn: that's exactly how it works
02:21:48 Quit lodesi_ ("leaving")
02:21:49preglowamiconn: for every byte in one of the multiplicands that is zero, you shave a cycle off the mul
02:21:58preglowvery annoying
02:28:34 Quit netmasta10bt (Read error: 104 (Connection reset by peer))
02:35:37 Join Sinbios [0] (n=Sinbios@Kingston-HSE-ppp3562416.sympatico.ca)
02:38:10 Join webguest70 [0] (n=4299238e@labb.contactor.se)
02:38:55webguest70Hey, I'd like to ask a question: I just installed rockbox on my nano and All the music is still in it's F00 type folders with random names anyway to get rockbox to load the Itunes DB so I can actually tell whats what?
02:41:53Paul_The_NerdRead the TagCache section of the manual
02:42:02Paul_The_NerdRockbox won't use the iTunes DB, but it can create its own
02:42:28webguest70Okay, Thank you
02:43:25 Join fatherfork [0] (n=4898d815@labb.contactor.se)
02:44:44 Quit Kohlrabi ("Quit")
02:44:55fatherforkhello everyone
02:47:23webguest70hi
02:49:01 Part pixelma
02:50:19 Quit webguest70 ("CGI:IRC (EOF)")
02:52:50Rudy4PezHey, is there any special reason Rockbox can't use the iTunes DB?
02:54:10Paul_The_NerdWe don't want to?
02:55:13 Join safetydan [0] (n=cbca159f@labb.contactor.se)
02:57:38Rudy4PezWhy not?
02:58:53Rudy4PezDon't get me wrong, I mean, I'm all for using the file structure to navigate.. Just like my Archos, but it'd make sense to have the option for TagCache to be able to index both files loaded through iTunes and files loaded normally all into one group, wouldn't you say?
03:00
03:00:01Paul_The_NerdTagCache indexes *all* files on the device
03:00:07Paul_The_NerdThat includes those synced over by iTunes
03:00:23Paul_The_NerdJust because it renames them something screwy doesn't mean TagCache ignores them. It's only interested in the metadata anyway.
03:00:34Paul_The_NerdWell, all supported music files on the device.
03:00:40Rudy4PezBut it's looking for ID3 tags... Do iTunes-loaded files carry that information?
03:00:47Paul_The_NerdIt's looking for metadata.
03:00:49safetydanRudy4Pez: yes
03:00:53Paul_The_NerdOn FLAC files it looks for Vorbis comments
03:00:59Rudy4PezAh, ok, sweet.
03:00:59Paul_The_NerdOn AAC files it looks for MP4 metadata.
03:01:17Paul_The_NerdNone of this has anything to do with the iTunes Database...
03:02:54Rudy4PezI'm sorry, I didn't know conversational tangents weren't permitted.
03:03:17 Join Criamos [0] (n=Criamos@p54933AD7.dip0.t-ipconnect.de)
03:03:17safetydaneh?
03:03:32Rudy4PezNothing.
03:03:35Paul_The_NerdI just meant that I don't understand how this ties into the iTunes database.
03:04:22Paul_The_NerdI wasn't complaining that the conversation went elsewhere. Just pointing out that you never brought up what about the iTunes database made you want it supported.
03:04:33Rudy4PezWell I was thinking of people who already had music loaded via iTunes, and how it could be made easily accessible in Rockbox.
03:04:49Rudy4PezFirst thing of which I thought, was to let Rockbox utilize the iTunes DB..
03:05:03Paul_The_NerdBut TagCache already covers finding and indexing the music.
03:05:31Rudy4PezBut you "don't want to" do that, so the second logical choice would be to let TagCache index it, which it apparently does..
03:05:34Rudy4PezAnd I didn't know that.
03:05:35Paul_The_NerdAnd it doesn't have filename restrictions, and can deal with music being added later by drag 'n drop.
03:05:55Paul_The_NerdI like the "s around the don't want to.
03:06:00Paul_The_NerdThere's no reason *to* want to.
03:06:11safetydanthere might be some scope for an "iTunes DB to TagCache DB" plugin. Converting things like smart playlists and so forth.
03:06:18Paul_The_NerdYeah
03:06:24safetydan(I'm assuming those things are stored in the itunes db on the iPod anyway)
03:06:30Paul_The_NerdThough the "smart playlists" still require iTunes to do the playlist updating, I hear.
03:06:43Paul_The_NerdThey only change at sync-time if i understand
03:06:56Paul_The_NerdNot as useful as the ones that could be created with a RuntimeDB search.
03:07:03 Join Febs [0] (n=medifebb@207-172-122-81.c3-0.rdl-ubr4.trpr-rdl.pa.cable.rcn.com)
03:07:14safetydanDoes it actually store the "playlist query" or does it just store the list of files that matches the query? i.e. does the iPod query the database, or just work through a list generated by the iTunes program?
03:07:21***Saving seen data "./dancer.seen"
03:08:06Paul_The_NerdI believe that the iPod just goes through a list
03:08:16Paul_The_NerdiTunes itself generates the list each time a sync happens, based on the query
03:08:18Paul_The_NerdOr so I've been told
03:08:23*Paul_The_Nerd doesn't use iTunes.
03:08:35safetydanheck I don't even use tagcache :)
03:08:53safetydantoo much effort to tag things
03:09:50BHSPitLappyI'd use tagcache
03:09:56BHSPitLappyif I had my iPod working at all...
03:10:04BHSPitLappysigh
03:10:19Paul_The_NerdI keep my files obsessively tagged, and obsessively filenamed, but don't use TagCache.
03:11:28Rudy4PezSame here.
03:11:54Rudy4PezHaving an Archos Jukebox forced me to keep a well organized music collection from the start.
03:39:17safetydanI wonder how people writing documentation can stay on task.
03:39:24*safetydan always gets distracted by code
03:40:53 Join daurn|laptop [0] (n=quae@124.243.137.107)
03:43:56 Quit Sinbios (Read error: 104 (Connection reset by peer))
03:49:20 Quit Criamos (Read error: 110 (Connection timed out))
03:51:06 Join |apo| [0] (n=r00t@dslb-084-057-064-105.pools.arcor-ip.net)
03:53:31 Join Davide-NYC [0] (n=chatzill@user-12hdtm5.cable.mindspring.com)
03:55:45JdGordonsafetydan: some ppl like writing code.. and some like wiritng docs... (which is more insane? :D )
03:56:03Paul_The_NerdClearly the docs people.
03:56:39 Join tommeyer [0] (n=icechat5@cpe-24-161-131-252.hawaii.res.rr.com)
03:56:44JdGordon:)
03:56:49BHSPitLappygatekeeper vs keymaster scenario
03:57:09sharpeeritten by tyhe code people.
03:57:09sharpeo document the code written by tyhe code people.
03:57:29JdGordonstay off the booze sharpe :D
03:59:26 Quit apo` (Read error: 145 (Connection timed out))
03:59:38 Quit sharpe (Read error: 54 (Connection reset by peer))
03:59:47safetydanIf someone could invent a text editor that does all the stuff that Eclipse lets me do to Java source to English "source" maybe things would be easier.
04:00
04:00:07JdGordonhmmm.. the software keylocking is annoyingly implemented :p
04:00:09safetydanEither that, or an AI that can generate user documentation from source code
04:01:40lamedamiconn: http://www.rockbox.org/tracker/task/996 I guess it was a long wait for that guy :)
04:02:12JdGordonlamed, can u do a commit for me while everyone is asleep? :D
04:02:42 Join sharpe [0] (i=sharpe@user-0c8hc2c.cable.mindspring.com)
04:02:52sharpestrange.
04:03:00JdGordonhttp://www.rockbox.org/tracker/task/5338
04:05:27safetydanJdGordon, how random is the folder selection? rand() % something usually isn't that random.
04:06:13JdGordonthat patch just is pretty random, it inserts the folder somehwere between two other folders in the current playlist.. its as good as its gonna get
04:07:43Davide-NYCanyone wanna help me with a patch?
04:08:05Davide-NYCdisktidy.patch is out of sync with the CVS by one entry
04:08:16safetydanJdGordon, do you have some objection to whitespace in code or something? :)
04:08:18Davide-NYCin other owrds I have to make one tiny change by hand
04:08:18safetydanint direction, first=playlist->started?playlist->index:0;
04:08:32Davide-NYCI DLed from CVS
04:08:33JdGordon:) its not that bad...
04:08:47Davide-NYCHow do I create a patch'
04:08:57JdGordonhad to remove spaces to stay under 80 chars p
04:09:00Davide-NYCI'm confused by this syntax from the wiki:
04:09:06safetydanI don't know about that, it's combining... five statements in one line
04:09:11Davide-NYCcvs diff -u file > patch
04:09:16safetydanJdGordon, that's what newlines are for
04:09:27Davide-NYCall changes are in apps/plugins/
04:09:49Davide-NYCI want to do a diff from CVS apps/plugins/ to a file called disktidy.patch
04:09:56Davide-NYChow do I proceed? :-)
04:10:27Davide-NYCAm I stupid? (please dont answer)
04:11:01FebsAre you trying to diff a single file or multiple files?
04:11:14Davide-NYCthe original patch affects three or four files
04:12:01Davide-NYCactually it affects three files (makes changes) and creates one.
04:12:12Davide-NYCHi Febs
04:12:23safetydanDavide-NYC: the command line you had seemed fine, just change "file" to "apps/plugins"
04:13:33Davide-NYCnot /cvsroot/rockbox/apps/ ?
04:13:49Davide-NYCor /rockbox/apps ?
04:13:56Davide-NYCand does it recurse?
04:13:59safetydanwell it depends on what your current directory is
04:14:02safetydanyes it will recurse
04:14:28Davide-NYCI have renamed rockbox-devel, will that break it?
04:14:36safetydanno
04:14:49Davide-NYChere goes (puts on helmet)
04:15:57lamedjdgordon: it's a bad idea if I'll put something into the source without it being discussed if it's a good or a bad thing to have
04:16:35JdGordon:) its been discussed a few times, noone will commit it tho, although i dont think there are acutlaly any objectinos to it
04:16:40lamedI might get slaughtered
04:16:52safetydanJdGordon, you might also want to clean up the code and fix the whitespace issues
04:17:06JdGordonclean up how?
04:17:34lamedI think it would be nice if you suggest it in the dev list
04:17:45Davide-NYCalright, the file I ended up with is missing the section that creates the disktidy.c file. Otherwise it looks good.
04:17:52Davide-NYCwhat did I do wrong?
04:18:04safetydanthe whitespace is my biggest issue with it on a quick skim
04:18:11JdGordonk
04:18:29Davide-NYCI could just creat the patch manually and ask people to test it
04:18:39Davide-NYCcut 'n' paste from here really
04:18:52safetydanDavide-NYC, ah you'll need to do "cvs diff -uN apps/plugins > patch"
04:19:09Davide-NYCwhat doe the N do?
04:19:14Davide-NYC(and thanks)
04:19:49safetydanJdGordon, also shouldn't "RANDOMLY" be "SHUFFLED">
04:19:52safetydan?
04:19:58safetydanthough I suppose random is different from shuffle
04:20:04safetydanDavide-NYC: include new files in the diff
04:20:17Davide-NYCsame deal!
04:20:18safetydananyway, lunchtime
04:20:21Davide-NYC(wierd)
04:20:26Davide-NYCthanks
04:20:58safetydanDavide-NYC, try here for some help http://www.rockbox.org/twiki/bin/view/Main/WorkingWithPatches#Submitting_A_Patch
04:21:05safetydanI think you'll need another step to add the file
04:21:19safetydanactually, sorry http://www.rockbox.org/twiki/bin/view/Main/WorkingWithPatches#Creating_A_Patch
04:21:29Davide-NYCok thanks
04:21:36lamedsafetydan: iirc you have to falsely direct diff to /dev/null, otherwise it won't really create a new file. I could be wrong!
04:21:48Davide-NYCthat's in the old patch
04:22:06JdGordondiff -u3 /dev/null /new/file > file.patch
04:22:26Davide-NYCwoa! what's that?
04:22:33JdGordonto crate the patch
04:22:40JdGordonreplying to lamed i guess
04:22:41Davide-NYCfromCVS?
04:23:21fatherforkhey Shadowdog, you there?
04:24:17barrywardellDavide-NYC: I would recommend "cvs diff -uN apps/plugins > patch" then "diff -u /dev/null apps/plugins/newfile >> patch"
04:24:33barrywardellreplace "newfile" with the filename of the new file
04:25:24barrywardellthat add
04:25:37barrywardelladds the new file to the end of the patch
04:26:00Davide-NYC>> == append?
04:26:08barrywardellyes
04:26:12Davide-NYCcool thanks
04:26:19Davide-NYCBRB
04:27:33Davide-NYCworked!
04:27:56fatherforkI have new new theme and wps I'd like some feedback on, so if anyone has a minute to look at it, pm me and i'll give you a screenshot
04:28:15JdGordonpaste a link to the screeny and youll get better response
04:28:16Davide-NYCthanks... but I can't believe it's a two command process if you create a file. Seems wierd to me, but hey, I'm a noob
04:28:31*JdGordon heads off to uni
04:28:37fatherforktrue
04:28:49fatherforkone sec
04:32:56safetydanDavide-NYC, the process in the links I posted isn't a two command (after the cvsdo add bit anyway)
04:34:03Davide-NYCahh
04:34:11Davide-NYCI was reading the bit AFTER that section
04:34:30Davide-NYCso first cvsdo add file then generate the patch
04:34:53Davide-NYCcvs diff -uN [files] > patch... where [files] can be a directory
04:35:03Davide-NYCand it'll recurse to all of your changes
04:35:07Davide-NYCcorrect?
04:35:38fatherforkalright, here's a screenshot, let me know what you think: fatherfork/dump060807-223013.jpg">http://i19.photobucket.com/albums/b185/fatherfork/dump060807-223013.jpg
04:35:49fatherforkalso, I need some suggestions for the progress bar
04:36:23safetydanDavide-NYC: yes
04:36:39Paul_The_NerdI'm always astonished how many people theme their WPSes after either the iPod interface, or Winamp, or Windows Media Player, or some other already existing interface.
04:36:48 Join netmasta10bt [0] (n=torment@pool-72-64-144-248.tampfl.fios.verizon.net)
04:36:59fatherforkwell, why mess with a good thing
04:37:06Davide-NYCpaul I agree
04:37:06fatherforkI just like the aqua interface
04:37:29Paul_The_NerdThat assumes that I think that any of those *are* good things
04:37:31Davide-NYCI just use the default, although every once ina while I'd like to see what's playing next.
04:37:53fatherforkme too
04:38:01Paul_The_NerdI took the default, stripped out the peakmeter, and replaced it with a basic grayscale image.
04:38:04Davide-NYCactually, this is a good place to ask. Is there a WPS == default + what's up next?
04:38:13Paul_The_NerdYou could make one pretty easily
04:38:22Davide-NYCbusy learning how to patch
04:38:26Davide-NYC:-P
04:38:26fatherforkso no go for mine?
04:38:27Paul_The_NerdThe code for the default WPS is in the Recorder WPS gallery
04:38:44Davide-NYCarghh, OK OK
04:38:59Paul_The_Nerdfatherfork: What do you mean No Go? I mean, I personally dislike it because I think iPod interface clones are silly, but why on Earth should anyone here's opinion matter anyway?
04:39:35fatherforkwell, if I'm going to upload it to the gallery, I think someone other than me should like it
04:39:37Paul_The_NerdDavide-NYC: :-P Well, nobody's made one yet like that. Though if you skim the H100 gallery, there's a lot of text-based ones (that could very easily be used on any player) so one of those ought to be pretty similar to what you want.
04:39:45fatherforkshould I just go ahead and upload it anyway?
04:39:55Davide-NYCI go check it out
04:41:10 Quit hannesd (Read error: 104 (Connection reset by peer))
04:41:41Davide-NYCYO YO !! Please test my first ever patch!
04:41:44Davide-NYChttp://www.rockbox.org/tracker/task/2596
04:42:09Davide-NYCI didn;t write it, I just fixed it. Pretty stupid but I'm excited
04:42:13Davide-NYC:-P
04:43:46 Join hannesd [0] (n=light@gate-hannes-tdsl.imos.net)
04:43:52safetydanLooks like you've generated the patch correctly too
04:44:39Davide-NYCyay
04:45:45fatherforkquick question about wps
04:46:28 Join scorche [0] (n=scorche@208-110-158-46.customer.csolutions.net)
04:46:31fatherforkis the vertical line spacing only determined by the font being used?
04:46:32Davide-NYCQuestion: What does it take to get a plugin commited to CVS?
04:46:53Davide-NYCI find Disktidy to be universal for all DAP that are UMSes
04:46:59Paul_The_Nerdfatherfork: Yes.
04:47:03scorcheback!
04:47:07Davide-NYCOSes leave detritis
04:47:09fatherforkthanks paul
04:47:15Paul_The_NerdDavide-NYC: I honestly don't know why that one hasn't been committed.
04:47:29Davide-NYCcan you commit that sucker?
04:48:01safetydanWell if no one beats me to it, I can always do it when I get home.
04:48:08Davide-NYCIt needs the linux detritis added (mention of gnome leving crap behind) but other than that it's good to go
04:48:19Davide-NYC*leaving
04:48:46Davide-NYCPlease do, I think Mac and Win32 is the majority of users (even rockbox users)
04:51:13safetydanQuite probably. I use Linux and Gnome so I can probably extend it to deal with the folders it sprinkles about the place as well.
04:51:49Davide-NYCThat would be truly sweet.
04:53:05*Paul_The_Nerd thinks Rockbox really needs a better resampler.
04:54:14 Quit Rudy4Pez ()
04:58:32Davide-NYCwonders what a resampler is and how to display italic font in IRC
04:59:28Paul_The_NerdThat was an action
04:59:34Paul_The_NerdYou type /me whatever
04:59:38ShadowdogMUDavide-NYC, not italics in IRC, only bold and underline
05:00
05:00:12*Davide-NYC haha I think I got it
05:00:24Paul_The_NerdI was just referring to what adjusts the sample rate of music before playback, for example, upsampling a 32khz file to 44.1 before playback. It sounded very bad, and resulted in people thinking that there's something wrong with our MP3 decoder
05:01:11FebsSounds like a job for preglow.
05:01:26safetydanI think a new resampler is somewhere on preglow's todo list...
05:01:32safetydanMaybe we need another preglow :)
05:01:56 Join Jungti1234 [0] (n=jungti12@125.188.3.64)
05:02:00Paul_The_NerdFebs: Remember this thread: http://www.misticriver.net/showthread.php?t=37144 <−−- The guy who said it only happens on podcasts just PMed me a sample file. It's almost definitely the resampler doing the it.
05:02:09*Paul_The_Nerd fires up the cloning vats.
05:03:29 Join Rob2222_ [0] (n=Miranda@ACADECBF.ipt.aol.com)
05:03:45Jungti1234hello
05:07:22***Saving seen data "./dancer.seen"
05:07:53 Quit Rob2222 (Read error: 145 (Connection timed out))
05:08:11 Quit barrywardell ()
05:08:46Jungti1234hmm strange
05:10:38lamedhi
05:10:51Jungti1234Why should language translation file be using Latin characters?
05:11:10lamedcome again?
05:11:23 Quit secleinteer (Connection timed out)
05:11:27Jungti1234http://www.rockbox.org/twiki/bin/view/Main/NonLatinLanguages
05:11:29lamedwhat languages do you speak?
05:11:31 Join secleinteer [0] (n=secleint@70.230.187.169)
05:11:33Jungti1234korean
05:11:52Jungti1234see.. 'for Translators'
05:11:57lamedand you don't get korean on the quickscreen right?
05:12:19Jungti1234yep, but I mean, why should language translation file be using Latin characters....
05:12:35Paul_The_NerdFor the quickscreen and some other screens, only latin characters are available
05:12:59lamedbecause it's better then having squares all around wouldn't you agree?
05:13:12Jungti1234Does rockbox's default font support Japanese?
05:13:28lamedyou already know the answer
05:13:44Jungti1234ah, ok
05:14:03lamedit's a tradeoff, you get to use the problematic screens using english, until everything uses the user selected font.
05:14:34lamedI wrote that wikipage. could you clarify it for me or tell me what there is to clarify?
05:14:49safetydanGetting user selected font to work on all screens is going to hurt.
05:14:58safetydanEspecially if it's something really big like the unifont
05:16:23Paul_The_Nerdlamed: You probably want to state that "On some screens only sysfont may be used currently. Because only latin characters are available to it, all sysfont strings must currently be latin, or they will display as non-character squares." or something similar
05:16:55Jungti1234where is 'http://www.rockbox.org/tshirt-contest/'?
05:18:41lamedthat was old
05:19:06Jungti1234Rockbox's default font doesn't support Japanese. no?
05:19:13Jungti1234http://www.rockbox.org/codes_new.png
05:19:24Jungti1234but, there is Japanese.
05:20:55lameduse it :)
05:26:28lamedlisten, Paul_the_nerd, jungti1234, it's 6:30 am, I can't think. try to understand whether http://www.rockbox.org/twiki/bin/view/Main/NonLatinLanguages makes sense or improve it :) jungti1234, I'd love if you make some corrections for the lang_sysfonts_ strings, will upload them soon
05:26:54lamedgoodnight
05:27:33 Quit lamed ("6:30 pm !")
05:27:33 Quit fatherfork ("CGI:IRC (EOF)")
05:32:11 Quit secleinteer (Read error: 60 (Operation timed out))
05:33:21 Join secleinteer [0] (n=secleint@70.230.187.169)
05:36:43 Join fejfighter [0] (n=jeffro21@d58-108-90-87.dsl.nsw.optusnet.com.au)
05:45:39 Quit Jungti1234 ()
05:48:24 Quit actionshrimp ("( www.nnscript.de :: NoNameScript 4.01 :: www.XLhost.de )")
06:00
06:03:41 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox 1.5.0.6/2006072814]")
06:06:03 Join Rudy4Pez [0] (n=A@65-102-20-210.ptld.qwest.net)
06:16:22 Quit midgey34 (Read error: 110 (Connection timed out))
06:32:03 Quit sharpe ("Leaving")
06:32:27 Quit secleinteer (Read error: 60 (Operation timed out))
06:33:08 Join secleinteer [0] (n=secleint@70.230.165.161)
06:33:36 Part Paul_The_Nerd
06:35:46 Quit rotator ("zzzzzzzzzz")
07:00
07:02:19 Quit secleinteer (Read error: 60 (Operation timed out))
07:03:03 Join secleinteer [0] (n=secleint@70.230.157.17)
07:07:24***Saving seen data "./dancer.seen"
07:15:47 Join jhMikeS [0] (n=jethead7@adsl-68-73-200-163.dsl.sfldmi.ameritech.net)
07:16:36 Quit _jhMikeS_ (Read error: 104 (Connection reset by peer))
07:20:14JdGordonsafetydan: still around?
07:29:01 Quit S0ap (Read error: 110 (Connection timed out))
07:30:10Bg3rmorning ;)
07:30:55JdGordonmorning
07:31:20 Quit secleinteer (Read error: 54 (Connection reset by peer))
07:33:13 Join secleinteer [0] (n=secleint@70.230.156.81)
07:36:37JdGordon*couph* task 5338 is synced again to cvs.... someone commit it...
07:40:36 Join spiorf [0] (n=spiorf@host69-214.pool8248.interbusiness.it)
07:48:18 Quit Rudy4Pez (Read error: 145 (Connection timed out))
07:54:06 Join EbErT [0] (n=EbErT@adsl-211-88-7.asm.bellsouth.net)
07:54:45ScoTTie_anyone ever replaced an iPod battery before ?
07:57:38JdGordondoes anyone else think show_main_backdrop(); should be defined for non colour lcds to remove the #ifdef HAVLE_LCD_COLOR suorungind it for every call?
08:00
08:00:05 Join theli_ua [0] (n=theli@mail.ukrcard.com.ua)
08:01:22 Quit spiorf ("Read error: 104 (Connection reset by peer)")
08:02:38 Quit secleinteer (Read error: 60 (Operation timed out))
08:03:24 Join secleinteer [0] (n=secleint@70.230.157.168)
08:08:12Bg3rScoTTie_: i'm sure there are many
08:08:40ScoTTie_im just wondering which sites they bought them from
08:08:57ScoTTie_the ones ive found look a little shady
08:09:59 Quit midkay ("Leaving")
08:11:03 Join midkay [0] (n=midkay@rockbox/developer/midkay)
08:20:57secleinteerhey has any1 had the problem that when you're charging and playing music at the same time, when the battery gets to 100%, rockbox freezes and you have to do a menu+select restart?
08:21:33Bg3rScoTTie_: ebay, hm...
08:29:52 Join Zagor [0] (n=bjst@rockbox/developer/Zagor)
08:33:11 Part EbErT
08:41:26safetydanJdGordon, yeah still around
08:51:02safetydanbut not for long
08:51:03safetydanhome time
08:51:05 Part safetydan
09:00
09:01:49 Join mikearthur [0] (n=mike@82-41-205-190.cable.ubr11.edin.blueyonder.co.uk)
09:04:48 Join ender` [0] (i=null@84.52.165.220)
09:04:53 Join petur [0] (n=d4efd6a6@labb.contactor.se)
09:07:26***Saving seen data "./dancer.seen"
09:07:59 Quit _Veseliq_ (Remote closed the connection)
09:17:05JdGordondoes anyone know how to open an ipod?
09:17:18Bagderhammer?
09:17:28*Bagder ducks
09:17:43JdGordonho ho ho :p
09:17:55Bg3r:D
09:17:58peturgoogle?
09:18:19JdGordonwaiting for ff to open
09:18:23scorche'lo Bagder
09:18:34Bagder'eeeeeello
09:21:23 Quit midkay ("Leaving")
09:21:44 Quit uski (Read error: 104 (Connection reset by peer))
09:21:58 Join midkay [0] (n=midkay@rockbox/developer/midkay)
09:24:39Lynx_Good morning! Is there a wiki page yet about video compression to work with the shiny new decoding features?
09:25:46scorchesomething happen over the weekend that i was not here? o_O
09:26:40markunLynx_: you can create one
09:27:03Mikachuscorche: mpeg player
09:27:24Lynx_markun: i might, once i find the mencoder command someone pasted here yesterday
09:27:30scorcheMikachu: go on... =)
09:27:38MikachuLynx_: you mean me?
09:27:48Lynx_Mikachu: might have been you, yes :)
09:27:54Mikachu(but it wasn't mencoder)
09:28:02Mikachuat least, mencoder encoded files don't work for me
09:28:31 Quit secleinteer (Read error: 60 (Operation timed out))
09:28:42markunLynx_: this one? 23.07.45 # <Bger> mencoder input_file -vf scale=224:176 -of rawvideo -ovc lavc -lavcopts vcodec=mpeg2video -o output.m2v <= this should work
09:29:15Mikachuit should, but it probably won't
09:29:19 Join secleinteer [0] (n=secleint@70.230.157.168)
09:29:30markunMikachu: any idea why not?
09:29:41Bg3rLynx_: yep, very probably this won't work ...
09:29:55Mikachumpeg2dec -o pgmpipe < input.m2v | pgmtoy4m -r 25:1 -a 4:3 | y4mscaler -O size=176x128 | mpeg2enc -a 2 -f 8 -q 6 -D 10 -E -10 -K tmpgenc -4 1 -2 1 -o output.m2v
09:30:03Lynx_ok. i'm just looking for a windows way of transcoding video to mpeg5
09:30:05Lynx_2
09:30:07Mikachuyou can use that to recode an already mpeg2 stream
09:30:23Mikachuno idea about windows
09:30:42Lynx_well, with video playback in cvs that question will crop up frequently i guess
09:31:04Mikachuyou can use these two commands for other files
09:31:19Mikachumplayer input.file -vo yuv4mpeg -vf scale=176:128 -ao null -ac null
09:31:26Mikachucat stream.yuv | mpeg2enc -F 4 -a 2 -f 8 -b 400 -4 1 -2 1 -Q 1 -D 10 -E -10 -K tmpgenc -4 1 -2 1 -o output.m2v
09:31:37JdGordonw00t... got the bitch open
09:31:39Mikachuyou can run mkfifo stream.yuv first to avoid the temporary file
09:32:14Mikachuand that is the sum of my knowledge on the matter
09:33:31Lynx_ok, i'll try to find the most simple windows way of transcoding
09:33:35Mikachu-b 400 is the bitrate and -F 4 sets the framerate, use -F 0 to get a list of possible values there
09:34:03JdGordonwhat should the ipod do if i plug my hdd from the h300 into it?
09:39:14peturgive some screen that it's borked and you need to run the repair tool?
09:41:12Bagderare they using the same kind of disk connector?
09:41:25JdGordonye
09:41:36JdGordonsame discs
09:41:53Mikachulinuxstb: you may want to reset the idle timer after the video finishes playing
09:44:19 Join barrywardell [0] (n=barrywar@host-194-46-234-58.dsl-ie.utvinternet.net)
09:47:36JdGordonhmm... fark.. why is my h300 not turning on after i opened t?
09:48:01Mikachua mystery indeed
09:48:17JdGordonturns on with AC, so still alove :p
09:50:25JdGordonwierd.. rb refuses to start?
09:50:54JdGordoni can turn it on with the ac plugged in, but if i try pressing on with no ac it just stays at a black screen
09:53:15 Quit secleinteer (Read error: 54 (Connection reset by peer))
09:55:06JdGordonhow do i do a disk scan/check in linux? i have a suspicion the disk is fubar..
09:55:18scorchefsck
09:55:37 Quit fejfighter ()
09:55:40BagderJdGordon: dosfsck
09:55:59JdGordonta
09:56:17 Quit mikearthur (Remote closed the connection)
09:56:35Lynx_w00t, video works. somewhat.
09:57:11JdGordonhmm.. all dosfsck tells me is that the free clusters are marked wrong.. how od i do a proper disk scan?
09:57:32Bagderwhat's a "proper disk scan" ?
09:57:34Lynx_ok, so Bgers mencoder command works (with -nosound added), and the video runs on my H300. But it often has strong block artifacts
09:57:40JdGordonsurface scan?
09:58:40Bagderno idea
09:58:50Lynx_umm, how would i stop video playback?
09:59:19 Join secleinteer [0] (n=secleint@70.230.145.7)
09:59:47JdGordonbadblocks for future reference
10:00
10:01:06Lynx_should pressing stop quit the video plugin? i seem to have to reset the player to stop it
10:02:02MikachuLynx_: yes, there's no exit button
10:02:09Lynx_Mikachu: ok
10:02:21MikachuLynx_: and yeah, that's what happened to my mencoder videos to
10:02:22Mikachuo
10:02:41Lynx_Mikachu: ah, so it's only mencoder
10:04:29JdGordonis there anyway to restore the ipod without the 56mb apple program?
10:04:51Mikachudo the rockbox install backwards
10:05:07JdGordonwould that work?
10:05:17Mikachumake sure the partitions are correct, then just write back your backups of the mbr and sda1
10:05:36Mikachuworked for me every time, never got their restore program to recognize the ipod
10:05:37JdGordonthey arnt... im testing a difeent hdd with the ipod
10:05:52Mikachuwell, set them up as they should be then
10:07:02 Join safetydan [0] (n=dan@cor4-ppp3002.bri.dsl.connect.net.au)
10:07:06JdGordondisk is doing alot of clicking which it doesnt do in my h300, so sounds like a hardware problem and not the hdd itself? would rockbox possibly solve that?
10:07:16JdGordonor wishful thinking?
10:07:27Mikachusounds unlikely to me
10:08:33JdGordonworth a try anyway.. :p
10:08:41 Quit sven_ (Read error: 113 (No route to host))
10:08:51JdGordonthe disk from the ipod seems to work fine in my h300
10:11:23 Join sven_ [0] (n=sven@port-212-202-193-252.dynamic.qsc.de)
10:17:15JdGordongrr... nope, installed rb and tryed booting but didnt work
10:17:47lexso... how does the mpeg player work? :p
10:17:57lexanyway, i'll try it now
10:20:23JdGordonok, maybe not!!!! it just booted
10:21:05*Bagder is stuck in a loop: patch, build, reflash, fail, patch, build, reflash, fail...
10:21:38*Mikachu gives Bagder a glass of pulverized birds
10:22:06Mikachu(that is only fun if you've played starship titanic)
10:22:36*Bagder is incompatible with games
10:22:45zeMikachu: that was a game?
10:23:01*petur has only read starship titanic
10:23:07*ze too
10:23:43peturhttp://www.starshiptitanic.com/
10:24:32Mikachuthere is also a game, by the company douglas adams started
10:24:41JdGordonhehe, backlight fading is awesome
10:24:43Mikachuhe is slightly featured in the intro and ending
10:24:55 Join dan_a [0] (n=dan-mirc@wormhole.domicilium.com)
10:26:05Mikachuit also features terry jones and john cleese as voice actors
10:26:43*petur discovers the game.... win95? yuck
10:26:45Bg3rBagder: patching/building/reflashing what ? the sansa devboard ?
10:26:53Mikachuheh, yeah it's old
10:27:00BagderBg3r: nah, my work board
10:27:09Bg3rah:)
10:27:11Bagderdoing a linux port
10:29:23*amiconn is thinking about viewports, and would like to hear opinions about a special problem
10:29:55amiconnToday, the scrolling code decides that it's time to drop the scrolling lines when the whole screen gets cleared
10:30:30amiconnWith viewports, we need a way to destroy a scrolling object (yuck, sounds like oo)
10:31:13amiconnI wonder what would be the best way to do this...
10:31:34JdGordoni think rockbox fixed this ipod!
10:31:38JdGordon3 cheers to all
10:32:36amiconnViewports will be structs provided by the app code that uses them. Usually the firmware layer does not need to store them, except if it comes to scrolling
10:32:43 Join qwm [0] (n=qwm@h229n1fls34o1010.telia.com)
10:33:42amiconnThe scrolling code will have a number of viewport "slots" for scrolling text objects
10:40:05JdGordonok, my h300 is refusing to turn on unless the ac is plugged in??? any ideas whats wrong? the batt is full
10:40:34peturdoes it power down when ac is removed?
10:41:08JdGordonit doesnt power down, it just turns off
10:41:17JdGordonlike a crash..
10:41:30peturbattery connected ok?
10:41:52JdGordoni unplugged it and plugged it back in.. i guess ill check again
10:42:55Bg3rJdGordon: i hope u haven't messed the polarity ?
10:43:26JdGordononly goes in 1 way
10:43:31 Quit slarti_ (Remote closed the connection)
10:45:41JdGordondoes ipod not charge from ac?
10:46:16JdGordonno, it does.. the icon isnt the same as on h300...
10:49:09 Quit midkay (Read error: 104 (Connection reset by peer))
10:49:36 Join midkay [0] (n=midkay@rockbox/developer/midkay)
10:50:25Lynx_did anyone try to compile mpeg2enc in cygwin?
10:52:10 Join _Veseliq_ [0] (n=veseliq@195.85.215.210)
10:54:26linuxstbLynx_: No, but afaik, it should work.
10:54:43Lynx_linuxstb: no, it doesn't here
10:55:00 Join slarti [0] (i=slarti@gentoo/developer/slarti)
10:55:05linuxstbWhat's the problem?
10:55:25Lynx_linuxstb: motionest.cc:780: error: invalid conversion from `unsigned int*' to `uint32_t*'
10:55:30JdGordonfuck, found the problem.. broke the battery wire :'(
10:55:59linuxstbLynx_: Are you using their latest CVS, or a released package?
10:56:22Lynx_linuxstb: i got the newest tarball from their site
10:56:26Bg3rJdGordon: better that than something else...
10:56:40JdGordonye. i guess so
10:57:17JdGordonlinuxstb: my friend owes u heaps... installed rockbox on the ipod seemed to fix its fucked up hdd!
10:58:09JdGordondinner!! bbs
10:58:40 Quit secleinteer (Read error: 60 (Operation timed out))
10:59:30 Join secleinteer [0] (n=secleint@70.230.178.138)
10:59:30linuxstbLynx_: If I was you I would give up and use one of the many MPEG encoders there are for Windows. tmpgenc seems highly regarded, but I'm not 100% if it's still free.
10:59:58Galoisif I were you I'd give up and install linux
11:00
11:00:31Lynx_linuxstb: yes, i tried mencoder and ffmpeg, but those give bad results on the player. i'll look for tmpgenc.
11:00:42Mikachuhttp://www.tmpgenc.net/en/e_download.html
11:01:05linuxstbMikachu: Did you ever try playing your bad mencoder encodes with mpegdec?
11:01:13linuxstb^mpeg2dec
11:01:29Mikachui don't think so
11:01:39Mikachudo you use their decoding code?
11:01:44linuxstbYes.
11:02:08linuxstbmpeg2dec is really just a demo program that comes with libmpeg2.
11:02:37linuxstb(but it's also very useful for feeding mpeg2enc)
11:02:55linuxstbIt has SDL and X11 output as well as output to files.
11:04:05Mikachulooks fine in mpeg2dec but not in mpegplayer
11:04:25linuxstbOK, I'll try and have a look. What resolution did you use?
11:04:42Mikachu176x128
11:04:47linuxstbAlso, what version of mpeg2dec do you have installed?
11:04:57Mikachu0.4.0
11:05:09Mikachumy libmpeg2 is 0.4.0b
11:05:10linuxstbOK - should be identical to mpegplayer then.
11:05:36linuxstbYes, 0.4.0b is what I used. But I've since discovered there have been a few changes in CVS, so I need to sync with those.
11:06:38lexargh
11:06:39 Join einhirn [0] (i=Miranda@bsod.rz.tu-clausthal.de)
11:06:43Mikachuthere are two patches in gentoo, altivec-fix-0.4.0b.diff gcc34-inline-fix-0.4.0b.diff
11:06:51lexlinuxstb: how can i convert my videos to mpeg-1 with a good resolution for my photo ipod?
11:07:03lexlinuxstb: on windows atm
11:07:12linuxstb1) Install Linux; 2) Use mpeg2enc....
11:07:26linuxstbSorry, I know nothing about video encoding under Windows.
11:07:28lexyeah, so i need to reinstall my grub only
11:07:30***Saving seen data "./dancer.seen"
11:07:37lexand then i'll get back to my amarok again too
11:08:14Mikachuffmpeg is a bit faster than mpeg2enc, encodes at 60fps
11:08:49lexi'll try with it...
11:09:40Mikachulex: or try tmpgenc, see url above
11:09:53linuxstbIs it free?
11:10:04lexMikachu: it whines about the resolution
11:10:13lexMikachu: because i changed it to 220x176
11:10:27Mikachubleh, wine decides to not show any fonts
11:10:33linuxstbMPEG video needs to be a multiple of 16x16 - so 224x176
11:10:39lexoh
11:10:40lex:)
11:10:56lexlet's try again then
11:11:10lexand 15fps or what?
11:12:21*amiconn should test whether the mpegplayer likes videos encoded with tmpegenc on windows, but that requires porting the mpegplayer plugin to h300 first...
11:12:30linuxstbOn the Photo, it manages around 13.5fps for a full-screen video. Reduce that down to Nano-size (176x128), and the fps goes up to about 24fps.
11:12:43linuxstbamiconn: No porting needed - it should work (but slower without lcd_yuv_blit)
11:12:52preglowhmm
11:12:55amiconnlinuxstb: Is it always 16x16, or only for colour?
11:12:57preglowhas it been optimised much?
11:13:15amiconnI think the macroblocks are 8x8, so the 16x16 requirement comes from the chroma subsampling
11:13:29 Join mkey [0] (n=mkey@pD9E354E7.dip.t-dialin.net)
11:13:40linuxstbpreglow: Not at all really. Apart from a "lcd_yuv_blit()" function I've added to combine the yuv2rgb conversion and writing the data to the LCD.
11:14:07linuxstbBut even that function is just a naive first implementation - that needs optimising.
11:14:31linuxstb(there's a version for the Color/Photo and Nano in CVS, and I've written one for the 5g which I haven't committed yet).
11:14:45preglowamiconn: macroblocks are 16x16, agaik
11:14:47preglowafaik
11:14:53Mikachupreglow: it was written yesterday
11:15:06preglowamiconn: transforms blocks are 8x8
11:15:29Mikachuhah, tmpgenc can't open m2v files
11:16:36peturrename to mpg
11:16:44 Join Arathis [0] (n=doerk@p508A6E92.dip.t-dialin.net)
11:17:40amiconnlinuxstb: lcd_yuv_blit() can only handle 2x2 subsampling... :(
11:18:15Lynx_linuxstb: tmpgenc is not free, at least not for mpeg2
11:18:19lexhttp://evot.us/x/rockbox/SSM11235.JPG hmm? :)
11:18:27lexis the problem in my compression or in rockbox?
11:19:19lexbut it show correctly on WINDOWS MEDIA PLAYER, so.. :o
11:19:46linuxstbamiconn: I know... If you think it will be useful elsewhere, I'll make it general again (assuming the performance hit isn't significant, which I doubt it will be).
11:21:10amiconnThe implemetation in xlcd is probably somewhat slow, because it uses division. If we restrict the subsampling to powers of 2 and use shifts, the performance hit should be minimal
11:22:00amiconnHmm, lcd_yuv_blit() won't be that useful for jpeg though
11:22:39amiconnIt won't work well with scrolling; would require redrawing the whole screen
11:23:08 Part astecp
11:23:11Mikachudoes the lcd controller have builtin support for scrolling?
11:23:51preglowso, now to decide how to handle audio :-)
11:24:09Mikachulibmpa can already handle mpeg1 and 2 audio?
11:24:27preglowwouldn't we pretty much want to use libmad?
11:24:33preglownot having to optimise stuff over again
11:24:45Mikachumaybe i meant mad
11:24:59preglowthe one we use in rockbox
11:25:05markunpreglow: or ac3?
11:25:30preglowmarkun: that's what i mean, it can contain anything (almost). ideally, we would want to use our ordinary codecs for this :)
11:27:01 Nick kclaf47 is now known as kclaf (i=kclaf@crj95-3-82-237-150-15.fbx.proxad.net)
11:27:02Mikachucan mpeg files really have ac3?
11:27:08preglowthat would mean having to solve the problem with iram, since the video codec would almost certainly also want to use that
11:27:18lexshould the mpegplayer work properly on a photo ipod?
11:27:19 Part kclaf
11:27:31lexwith mpeg-2 i'm getting a bit better image quality but this sucks too
11:27:47lexhttp://evot.us/x/rockbox/SSM11235.JPG and that was a mpeg1 with sound
11:27:48markunMikachu: we can store the m2v and the audio stream in a avi container
11:28:35 Quit secleinteer (Operation timed out)
11:28:43markunlex: I think you need to extract the video stream from your mpeg file
11:29:11merbananMikachu: yes, but I'm not sure it is standardized
11:29:19 Join secleinteer [0] (n=secleint@70.230.155.44)
11:29:19lexmarkun: how does that happen on windows? :o
11:29:30 Join crashd [0] (i=nobody@badger.ing.me.uk)
11:29:49crashdmorning all
11:30:47linuxstbpreglow: On the ipods, it seems that one CPU will be enough for full-speed video playback (and bitting to the LCD) - the Nano is 24fps already. So my plan is to use the COP for video, and the main CPU for disk buffering and audio decoding. I don't think the cache will be an issue there - the video thread should hopefully be reading from a part of the ringbuffer that has long since been removed from the cache.
11:31:09markunlex: no idea. with graphedit maybe
11:31:38lexmarkun: i converted my .avi to .mpg as mpeg-2 with no sound
11:31:48lexshould it be .mpeg? no difference? :p
11:32:01safetydanC99 still isn't welcome even in plugins right?
11:32:25lexhttp://evot.us/x/rockbox/SSM11237.JPG it looks like that with mpeg-2
11:32:45lexbut that's a good start :)
11:32:55lexdo you have a good working mpeg for me? :p
11:34:31markunlex: search for a program to 'demux' a mpeg file for windows. Then extract the video stream from your file
11:34:48Mikachulex: i can send a file for nano, should work on photo i guess but won't fill the scren
11:34:50lexmarkun: no, i'll reinstall my grub and go to linux
11:35:01lexMikachu: yes, send it, please :)
11:35:07lexMikachu: it's better than nothing
11:35:24lexmarkun: and then i'll get into amarok again
11:35:38lexnice speed :)
11:35:43lex(12:37:18) [freenode] DCC received file ambitious-176x128.m2v [11MB] from Mikachu in 00:00:10 [1196.11kB/s]
11:35:59Mikachuheh
11:36:03Mikachutoo bad i don't have 100mbit anymore
11:36:31lexi have only 10/10 on this server
11:36:36lexand a dedicated 100/100 on the other
11:36:41lexand a 2/512 at home :D
11:36:50lexthis really sucks
11:36:58lex220KB/s <3
11:38:18lexthat video works fine
11:38:21lexthanks :)
11:38:30lex27.6 fps
11:38:42Mikachuit is faster in the start for some reason
11:38:49Mikachu28, then drops to 24.5
11:38:55Mikachuat least on my nano
11:39:21lexme too
11:39:27lex23.3 atm
11:40:29lexbut how can i quit from this player
11:40:36lexhttp://evot.us/x/SSM11238.JPG yay
11:40:52Mikachuyou have to wait
11:40:55Mikachuor reset
11:41:08linuxstbMikachu: I would welcome a patch for an exit button :)
11:41:10lex:(
11:41:20Mikachulinuxstb: i thought you left button code out for performance reasons
11:41:21preglowsafetydan: i don't see why it shouldn't be
11:41:30lexsorry, http://evot.us/x/rockbox/SSM11238.JPG
11:41:50preglowlolrofl
11:41:54preglowsomeone send me a test file!
11:42:04Mikachui can send you the same file
11:42:10preglowgo ahead
11:42:22linuxstbMikachu: No, just laziness.
11:42:32safetydanpreglow, well... maybe I'm misremembering... I'm sure there's been some objection to C99 constructs before
11:42:36Mikachu3MB/s is pretty good for 10mbit
11:42:37lexwait
11:42:40preglowMikachu: thanks
11:42:41lexi have uploaded it
11:42:45lexah, never mind :)
11:42:50lexhttp://evot.us/x/ambitious-176x128.m2v it's there
11:42:55preglowMikachu: my irc shell is on 100mbit
11:43:05Mikachuyes, but my connection isn't
11:43:18preglowsafetydan: all we use is gcc anyway
11:43:23lexMikachu: how did you convert this?
11:43:25Mikachuwell, the ethernet is 100mbit, but my account is supposed to be 10mbit
11:43:34amiconnsafetydan: _some_ C99 features are discouraged, like mixing variable declarations with code, and // comments
11:44:17preglowwhy?
11:44:35amiconnReadability
11:44:49Mikachulex: http://mikachu.ath.cx/rockbox/mpeg2encoding.txt
11:45:02lex:)
11:45:08preglowsome code gets more readable if you mix variable declarations and code, if you ask me
11:45:10lexi'll go to reinstall my grub now
11:45:17safetydanamiconn, those are what I was talking about. But are they as much of a problem in plugins?
11:45:25preglowas for comments, i don't get why // are less readable than /* */
11:45:32lexMikachu: so, if i want to convert videos for my photo, i'll change size= to size=220x176? :)
11:45:43lexMikachu: or scale
11:45:44linuxstb224x176...
11:45:49lexyeah, sorry
11:45:55Mikachuyes
11:46:05lexso, reboot ->
11:46:21 Quit merbanan (Read error: 54 (Connection reset by peer))
11:46:28linuxstbOr 224x128 for 16:9 videos.
11:46:41 Join dionoea_work [0] (n=3ea147f9@poy.chewa.net)
11:46:45dionoea_workhello
11:46:45amiconnsafetydan: They are less of a problem in plugins, but still better avoid them for consistent style
11:47:06 Quit petur ("worrrk")
11:47:07 Join merbanan [0] (i=banan@farmer.campus.luth.se)
11:47:17preglowthis reeks of jpop!
11:47:38dionoea_workMikachu: ac3 is standard with mpeg2 (think DVDs)
11:48:05linuxstbdionoea: _a_ standard... MP2 audio is very common as well (think DVB)
11:48:08safetydanlooks like I've got some cleanup ahead of me
11:48:11JdGordon:'( i killed my h300 somehow.. or just the battery.. its all soldered correclty and it still wont turn on :'(
11:48:22 Join Hansmaulwurf [0] (n=maerlyn@p50819EC8.dip0.t-ipconnect.de)
11:48:22amiconnIf I would know how to tell the mpeg2 decoder to not decode chroma, that might be useful...
11:48:29 Join lodesi [0] (n=lds@d01m-195-36-140-244.d4.club-internet.fr)
11:48:31merbanandionoea_work: it is standardized in vob I'm not sure it is in mpeg
11:48:51dionoea_worklinuxstb: well AFAIK ac3 also works in DVB ... although people usualy put plain mpeg audio part II
11:49:12linuxstbdionoea_work: Yes, it is. AC3 is part of the DVB spec.
11:49:27linuxstbBut at least in Europe, MP2 is by far the most common.
11:49:47qwmlinuxstb: the mpeg player, great work! plays mpeg2 at 25 fps smoothly on my nano. :) *very happy and optimistic*
11:50:19Mikachua lot of dvd files also have raw pcm in them
11:50:24Mikachuer, dvb
11:50:46linuxstbYes, it's ironic that the device least suited to video (smallest screen, smallest storage) is perfoming the best... But not unexpected.
11:50:56preglowdoes it decode video that is larger than the screen?
11:51:07linuxstbYes, it should crop it.
11:51:12JdGordonWOOOOTTT its ALLIIIVVVEE!!!
11:51:24Bg3rJdGordon: your h300 ?
11:51:30JdGordonye
11:51:40Bg3rnice to hear
11:51:41linuxstbpreglow: But it decodes to the full resolution (as YUV). The blit will crop and only convert to RGB what is displayed.
11:51:46qwmhehe. :)
11:51:54JdGordonfor some reason plugging in the ac and removing it got it running of the batt again (whch did actually need resoldering
11:52:03JdGordonso now i have a working 4g ipoo and h300 :D
11:52:18lexon a ubuntu cd now :o
11:52:28preglowlinuxstb: don
11:52:28amiconnDid someone already test video playback on coldfire (h300/x5)?
11:52:36preglowlinuxstb: don't think there's anyway not to decode everything
11:52:48preglowany way, even
11:52:53Bg3ramiconn: iirc about 9fps
11:53:18linuxstbpreglow: No, I don't think so.
11:53:22Mikachupreglow: in libavcodec you can specify lowres=X to decode only every xth pixel, but i don't know about precropping
11:53:27merbananpreglow: well you could skip some macroblocks
11:53:47Mikachuit makes some superhd files playable for me, and halves x and y resolutions
11:53:49merbananpreglow: won't be pretty though :)
11:54:09amiconnMikachu: I guess it works the same as jpeg downscaling (what our jpeg viewer does)
11:54:19Mikachuprobably
11:55:39dionoea_workyou might need those macroblocks to decode the next frames (i'm not too confident about my knowledge of mpeg2 decoding ... but i think that this is a safe bet)
11:56:13dionoea_workso you won't really be able to skip them ... or like only decode I frames
11:56:27JdGordonfark!! bloody wire came unsoldered again :'(
11:56:41dionoea_workput some glue :)
11:56:57Bg3rJdGordon: show you solder skills :D
11:56:57amiconnlinuxstb: What is the performance increase due to lcd_yuv_blit() on ipod?
11:57:10merbanandionoea_work: well if a motion vector points from the skipped blocks to the active screen thenit won't work
11:57:14Mikachui got about 3 or 4 fps better on my nano, iirc
11:57:22*amiconn guesses lcd_yuv_blit() would be even more useful on coldfire than on pp
11:57:24JdGordonBg3r: no, its my roughness trying to get the damn thing back in the case
11:57:37linuxstbamiconn: Definitely.
11:57:58merbananbut the video should be encoded to perfectly suit the player
11:58:14*linuxstb is also waiting for earHertz to submit his ARM memcpy to the patch tracker...
11:58:39 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
11:59:20 Quit amiconn (Nick collision from services.)
11:59:20 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
12:00
12:00:29crashdmorning
12:01:09dionoea_workdoes rockbox already handles audio stream demuxing seperately from the decoding ? or are you going to need to split that for the video decoder ?
12:01:42linuxstbThere is nothing written apart from playback of a demuxed video stream.
12:01:45 Join Kohlrabi [0] (n=Kohlrabi@dslb-082-083-143-245.pools.arcor-ip.net)
12:01:50dionoea_work(using the ogg demux wouldn't require too much porting work i guess)
12:02:01crashdwhere did all this video tlak come from?
12:02:02dionoea_worklinuxstb: i meant when you handle the audio in "normal" rockbox
12:02:17linuxstbI've been told AVI is the most efficient container for video, so that's probably what will get supported.
12:02:31linuxstb(in terms of being easy for a decoder to work with)
12:03:07linuxstbdionoea_work: No, the decoders handle the container format as well for audio.
12:03:08dionoea_workAVI is the worst container there is </personal point of view shared by many others> :)
12:03:19qwmfor how long have you worked on the mpeg player, linuxstb ?
12:03:21linuxstb"worst" = "simple" ?
12:03:21dionoea_workbut it's also the most used
12:03:34dionoea_worksimple == mpeg ts
12:03:52linuxstbqwm: I did most of the work on Sunday, plus some cleaning up yesterday.
12:04:05dionoea_workbut mpeg ts isn't really suited if you want to seek
12:04:13qwmamazing.
12:04:16linuxstbdionoea: mpeg-ts isn't the best - everything is in 188 byte packets, so lots and lots of memcpy
12:04:42linuxstbNone of the mpeg formats are suitable for seeking - they are streaming formats.
12:04:51linuxstb(apart from VOB which adds the NAV packets)
12:04:52merbananlinuxstb: you might think of using not as container
12:05:00merbananlinuxstb: er, nut
12:05:16amiconnlinuxstb: Does the mpegplayer already use iram?
12:05:46linuxstbmerbanan: OK, I'll look at it. But that's a long way away - I'm planning on just reading separate .mp2/.mp3 files for the first tests with audio playback.
12:06:03linuxstbamiconn: Very little.
12:06:17merbananlinuxstb: if you only think of using a format especially for rockbox, then I think nut would fill the requirements quite well
12:06:25linuxstbamiconn: (and what is there is just randomly chosen)
12:06:25*amiconn is going to try video in the sim
12:06:36merbananlinuxstb: efficient, not floats and easy to seek in
12:07:01linuxstbmerbanan: It would be nice to use a standard format if possible - gives users a wider range of encoders to use, and means we don't need to write an encoder...
12:07:06amiconnmerbanan: We will also need an easy to use encoder tool then
12:07:33amiconnFor our .rvf format we use on archos, there are 2 ways to create such files.
12:08:56amiconn(1) A set of commandline tools for those who want full control, but this way is cumbersome
12:09:19linuxstbWhat bitrate would a .rvf file need to be for the H140?
12:09:24amiconn(2) A directshow filter for windows. This isn't gpl though
12:09:45linuxstb(just the video portion)
12:10:38merbananamiconn: true, well it's worth looking into if one aim for efficiency, but m2v would be most easy to use
12:11:19amiconn160 * 128 * 70 (lcd frames per second), i.e. roughly 1.4MBit/s
12:11:23linuxstbdionoea_work: Going back to AVI, what are the bad points about it, and do they affect Rockbox?
12:11:28amiconnSame bitrate as CD quality WAV
12:12:31linuxstbNot completely unrealistic then...
12:13:32dionoea_workno support for B frames, no support for subs, bloated format with no clear standard
12:14:09amiconnFor iPod g3/g4 it would be ~1.6MBit/s (higher lcd frame rate), for mini ~1.4MBit/s
12:14:12dionoea_workthat probably doesn't matter much for rockbox though
12:14:33JdGordonall together and working
12:14:41*JdGordon has no intention of openig his h300 any time soon :p
12:15:15amiconnlinuxstb: Hmm, shouldn't the m2v show up in the file browser?
12:15:21dionoea_workmp4 / mov might be better (especially if you ever need to encapsulate stuff like mp4/h264/dream on...)
12:16:04linuxstbamiconn: They do for me...
12:16:46linuxstbdionoea_work: I like the sound of mp4... Very well suited for Rockbox.
12:16:54dionoea_workmp4 is a container ...
12:17:18linuxstbI know, that's what I'm talking about.
12:17:23amiconnHehe, the H300 sim achieves >110 fps with the test video. Looks rather funny...
12:17:35crashdid vote for mp4
12:17:43dionoea_worklinuxstb: why wouldn't it be suited ? (i really don't know)
12:17:48crashddue to most of the stuff i have encoded for my ipod is in mp4 ;)
12:17:55linuxstbdionoea: I said "Very well suited".
12:17:57compubombhttp://trogdoor.googlepages.com/jesusfuckingchrist.jpg
12:17:58Bg3rwhy not simple avi ?
12:18:07dionoea_workoh, i though it was ironic
12:18:40dionoea_workcompubomb: is that link worksafe ? :)
12:18:40linuxstbBg3r: Nothing's ruled out yet. But AVI doesn't seem the best thought out container format.
12:18:51crashdhttp://en.wikipedia.org/wiki/Comparison_of_container_formats
12:18:53crashdnot great
12:19:04 Nick mkey is now known as Janina14 (n=mkey@pD9E354E7.dip.t-dialin.net)
12:19:04crashdbut it's a start if you're considering containers, altho, im sure linuxstb and co. alreayd know most of the stuff
12:19:05dionoea_workyou could allow several container formats ... but that might add computing overhead
12:19:13crashdright guys, i'm off to china for 10 days, good luck :)
12:19:15midkaydionoea_work: haha, it isn't :)
12:19:24 Nick Janina14 is now known as mkey (n=mkey@pD9E354E7.dip.t-dialin.net)
12:19:54Bg3rmatroska ?
12:20:28amiconnlinuxstb: Strange... m2v is in viewers.config for the sim, yet m2v files don't show up as supported. 'Open with..' works though
12:20:29dionoea_workmore apps support mp4 muxing than mkv muxing
12:20:48*amiconn is building for h300 target
12:27:01dionoea_workBg3r: AFAIK, mkv is also a bit more complicated to read than mp4
12:27:49Bg3rdionoea_work: okay, my suggestion was just a shoot in the dark
12:28:32dionoea_workbut it's true that mkv can do better stuff (like handle chapters)
12:31:26 Join mkey_ [0] (n=mkey@pD9E36E5F.dip.t-dialin.net)
12:31:37amiconnlinuxstb: Will you implement lcd_yuv_blit() for the g5 ?
12:31:51 Quit Kohlrabi (Nick collision from services.)
12:31:52linuxstbI have done, but haven't committed it yet.
12:32:02 Join Kohlrabi [0] (n=Kohlrabi@dslb-082-083-129-027.pools.arcor-ip.net)
12:32:03compubombdionoea_work: no
12:32:38linuxstbamiconn: Are you happy to leave it as it is, or do you think we should add the csub_x/y parameters back?
12:33:04amiconnI think we don't need variable csub* for now
12:33:21compubombbtw, i noticed that rockbox uses substantially more battery than Mac's ipod OS
12:33:27amiconn(but perhaps I'll change lcd_yuv_bitmap a bit)
12:33:50preglowcompubomb: congratulations
12:34:27compubombpreglow: is that sarcasm ?
12:34:31Mikachui do believe it is
12:34:49JdGordondoesnt the ipod stay in rockbox for usb?
12:35:11dionoea_workno
12:35:25dionoea_workunless you only want to charge the batteries
12:35:44crashdoh, one question before i go, what's the deal with keeping retailos running whilst charging?
12:35:44JdGordonis there a reason? or just no need?
12:35:58dionoea_worki think that it isn't implemented yet
12:36:26amiconnJdGordon: If you implement an usb stack and figure out the pp usb hardware, we could implement it...
12:36:26dionoea_workbut i'm not sure if that's the real reason :)
12:36:30amiconn</sarcasm>
12:36:40linuxstbcrashd: What do you mean?
12:36:47dionoea_workdoesn't ipodlinux already use usb ?
12:36:51amiconnusb is implemented in software on ipod
12:36:54linuxstbdionoea_work: No.
12:37:00crashdlinuxstb: like, in rockbox you can hold down menukey and charge, and still use rockbox yeah?
12:37:03JdGordonhmm.. crap, me has no usb then with the ipod
12:37:19dionoea_workwell you have the disk mode thing
12:37:20linuxstbcrashd: Yes. So you're asking about how to do the same in retailos?
12:37:34crashdlinuxstb: yeah, because i dont really use retailos, but ive got some vids i wanna watch whilst im on holiday
12:37:39crashdand i cant remember if you can even do it in retail
12:37:50dan_acompubomb: there are a lot of things we don't know about the ipod yet, for example how to turn some of the chips off when they're not being used.
12:37:53JdGordondionoea_work: no, disk mode is broken.. for some reason the retail os cant access the hdd properly
12:38:01dionoea_workouch
12:38:13 Join afrika[a] [0] (n=murderin@cm117.omega170.maxonline.com.sg)
12:38:14JdGordonye
12:38:18linuxstbcrashd: I've no idea... retailos has long since departed my ipods.
12:38:32crashdlinuxstb: exactly O.o
12:38:36 Part afrika[a]
12:38:56JdGordondoes rb play aac?
12:39:05linuxstbJust about...
12:39:19linuxstbThe decoder is slow, and the mp4 container parser is very fussy about the input stream.
12:39:29JdGordonk
12:39:36linuxstbBut I think 160kbps files will play on an ipod. 128kbps definitely does.
12:40:03linuxstb(Coldfire targets skip even with 128kbps files).
12:40:06compubombdan_a: so it's pretty much a crap shoot untill everything is reverse engineered etc.
12:40:13 Join PaulJam [0] (n=pauljam@p54BCDB55.dip.t-dialin.net)
12:42:38*amiconn found out why the .m2v didn't show up. Not rockbox' fault...
12:43:36*dionoea_work hopes that reverse engineering USB on portal player chips will be easier with the sansa jtag thing
12:44:07Mikachuyou'll also have to hope that someone wants to do it :)
12:44:26dionoea_workhehe
12:44:49amiconnlinuxstb: Something is wrong with mpegplayer on H300. It just shows "Buffering..." then hangs...
12:45:01JdGordonok, wierd... it connects fine with windows, but trygint to connect to linux goes straight to the "ok to disconnect" screen
12:45:57amiconnEhum, the usual voice problem :(
12:46:03 Quit mkey (Read error: 104 (Connection reset by peer))
12:46:06barrywardelllinuxstb: slight change of topic, but you mentioned before that you had a look at the H10 bootloader. Was the .hex or the .rom version?
12:46:36dionoea_workJdGordon: last time i had that kind of problem a rmmod ehci_hcd fixed it ...
12:46:49*barrywardell is trying to teach himself arm and figure out the bootloader
12:46:52linuxstbI think the .hex.
12:47:18amiconn15fps on H300... (with the 176x128 test video)
12:47:21linuxstbamiconn: Why does voice break it? I thought that was a menu problem.
12:47:39JdGordondionoea_work: thanks
12:47:48JdGordonseems to be working now
12:48:03 Join pascals [0] (n=pascal@ip503c8584.speed.planet.nl)
12:48:17pascalsHi Everyone
12:48:28pascalsAny ipod developer around?
12:48:43linuxstbDepends on your next question...
12:48:46pascals:)
12:48:52pascalsAlways, doesn't it :)
12:49:09pascals1 question, 2 remarks.
12:49:16pascals1) I like it !!!!
12:49:20amiconnlinuxstb: Not sure... perhaps because it's a viewer. I had file & directory voicing enabled, and maybe the voice was still babbling when the plugin claimed iram
12:49:22pascals(that out of the way)
12:49:55amiconnA stop button would be really useful... (hint hint)
12:50:07pascals2: The default WPS screen has a peak level meter, which makes my 4G/color skip sound. with 192/256kbit MP3s
12:50:08linuxstb:)
12:50:32pascalsAll the non-default WPS screens don't have this problem...
12:51:18amiconnFor some reason the frame rate goes down a bit while playing... ~13 fps now
12:51:26safetydanpascals, yeah the iPod code isn't quite fast enough to do peak meters and playing audio
12:51:42pascalsI admit the peak level meter is cool, but not having it and playing ok sounds like a better idea.
12:51:47barrywardelllinuxstb: one more quick question. what does writing 0x20 to GPIOF_ENABLE do? Does it enable bit 5 of GPIOF?
12:51:57linuxstbamiconn: Maybe that's a flaw in my fps calculations (rounding?) - pacbox does the same....
12:52:07pascalssafetydan: does the code allow for a different default WPS for iPods, then?
12:52:18amiconnlinuxstb: In the sim it goes up, especially towards the end...
12:52:28*pascals found the installation quite painless, btw.
12:52:32Febslinuxstb, where should that forum thread about video developments be moved? Rockbox General Forum?
12:52:49safetydanpascals, just pick one that you like instead of the default
12:52:51linuxstbFebs: Yes, I think so. It applies to all colour targets - ipods, X5 and H300.
12:53:30PaulJamweren't there plans to make the icatcher wps the default one?
12:53:31pascalssafetydan: Well, yes, but what I mean is - if you want more people to pick up rockbox, disable the peaklevel meter
12:53:42linuxstbbarrywardell: I'm not 100% sure...
12:53:48safetydanpascals, the problem will eventually go away as the code gets better/faster
12:53:57amiconnlinuxstb: YOu need to do sth about the activity thing. My H300 just shut down during m2v playback...
12:54:03Mikachui wouldn't want the peakmeter even if it didn't cause skipping
12:54:40amiconnI would rather use the rvf icon (film strip) for m2v instead of the double note
12:55:02 Nick mkey_ is now known as mkey (n=mkey@pD9E36E5F.dip.t-dialin.net)
12:55:03linuxstbpascals: Rockbox needs a better default WPS in general, not just removing the peakmeter. It will happen, but no-one likes temporary solutions such as commenting out the peakmeter in the default WPS just for the ipods.
12:55:20linuxstbamiconn: Yes, I was looking for a better icon. I'll swap it.
12:55:25*amiconn changes
12:55:31linuxstbPlease commit.
12:55:52pascalslinuxstb: perhaps a #ifdef for the target, then.
12:55:59daurn|laptopyo
12:56:04linuxstbpascals: That's what I mean by temporary solutions.
12:56:05pascalsStuff that doesn't work tends to keep people away.
12:56:16lexmy kernel sucks, it wont recognize my usb devices
12:56:28Mikachu#linux
12:56:30pascalslinuxstb: ok. :)
12:56:44lexMikachu: ... :(
12:56:52lexMikachu: #ubuntu.fi is better - on quakenet
12:57:02*pascals will be back in 30 minutes after fixing his car
12:57:04dionoea_worklex: you might need to load the appropriate kernel modules
12:57:22lexdionoea_work: ubuntu should do it automatically, but i'm installing a bit older kernel now
12:57:38dionoea_workdid you try sudo modprobe uhci_hcd ?
12:58:44 Part compubomb
12:58:47safetydanbarrywardell, is that OS X sim patch okay to commit? It doesn't seem to break anything for me under Linux, and I assume it works in OS X.
12:59:17JdGordonsafetydan: i cleaned up and synced the insert folders patch if your itching to commit something :p
13:00
13:00:31barrywardellit works in OSX. The only thing I'm not sure about is the change to apps/Makefile
13:01:32barrywardellit removes "-Wl,-Map,$(OBJDIR)/rockbox.map" since -Map isn't supported on OSX
13:01:44daurn|laptopfine
13:01:49daurn|laptopdon't respond to me
13:01:51daurn|laptopbastards
13:02:22safetydandaurn|laptop, did you ask a question?
13:02:24linuxstbsafetydan: Why does Disktidy not work on the player?
13:03:04barrywardellsafetydan: it also disables the bubbles plugin since that doesn't compile
13:03:15daurn|laptopi said "yo"
13:03:16daurn|laptop:P
13:03:41FebsPerhaps when you say something that calls for a response, you'll get one.
13:03:54safetydanlinuxstb, the original submission didn't support it
13:04:13safetydanbarrywardell, odd that bubbles won't work on one type of OS
13:04:36linuxstbbarrywardell: Why doesn't bubbles compile? Is it a bitmap issue?
13:04:54barrywardelli'll try compile it and get the error for you now
13:05:27linuxstbIt would be nicer to fix bubbles. It shouldn't do anything uniquely differently to any other plugin.
13:05:50safetydanJdGordon, not much time left in the night. If I get a chance, I'll look again.
13:06:13JdGordon:)
13:06:42barrywardellit's a linker problem: "undefined symbols: _bubbles_bubble"
13:06:50amiconnbarrywardell: -Map is disabled for OSX only?
13:07:18barrywardellamiconn: no, right now the patch removes the -Map part altogether
13:07:19safetydanamiconn, not in the patch as it stands
13:07:34***Saving seen data "./dancer.seen"
13:07:39linuxstbYou would get lots of complaints if you removed the .map file...
13:07:42amiconnThat's bad. The .map files are a very useful resource for debugging.
13:08:05*amiconn wonders why -Map isn't supported on OSX
13:08:07Mikachuwhy would they disable that feature on osx?
13:08:08barrywardell-M does work though but outputs the map to standard output
13:08:14Mikachudon't you use gcc there?
13:08:19barrywardellyup gcc
13:08:27barrywardell4.0.1
13:08:35linuxstbBut patched by Apple IIRC.
13:08:36amiconnMikachu: -Map is a linker option
13:08:49Mikachuer, binutils then?
13:09:02amiconnyes
13:09:39linuxstbMaybe they use some BSD tools, rather than the GNU ones...
13:09:51Mikachuyou can just require gnu ld then
13:09:51safetydanWhile I'm poking around Makefiles... should I remove the X11 and W32 sims and related Makefile entries?
13:10:58linuxstbbarrywardell: bubbles_bubble is an external bitmap. Do all the other plugins using bitmaps work?
13:11:36barrywardellbubbles.c is the only plugin I have come across that doesn't compile
13:11:48barrywardellwhat other plugins use bitmaps that I can try?
13:12:37Mikachubarrywardell: what does ld −−version say?
13:13:47barrywardell"unknown flag: −−version", ld -v says "Apple Computer, Inc. version cctools-590.42.1.obj~1"
13:14:05theli_uaanyone can say how hard would it be to port some code from i386 asm to arm asm ?
13:14:17amiconnlinuxstb: lcd_yuv_blit() does too many safety checks and conversions, imho
13:14:37amiconnIt's meant to be a performance function, so all that swapping/cropping can go away
13:15:12linuxstbamiconn: As I said, it's just a naive implementation to start with.
13:16:10linuxstbtheli_ua: You can't really port asm - you need to rewrite it...
13:16:31 Join jaczehack [0] (n=jac@c-d0f772d5.013-1-73746f40.cust.bredbandsbolaget.se)
13:16:53theli_ualinuxstb, well i mean how much differences there are ?
13:17:05amiconnEverything is different
13:17:09Mikachui don't think it would be unreasonable to require gnu ld, most people already have compiled it for the crosscompiling anyway
13:17:36Mikachui know you can't use the same binaries, but they wouldn't be very afraid of compiling it for mac native
13:17:37linuxstbamiconn: Although some cropping is needed - for 220x176 displays, the video needs to be encoded at 224x176.
13:18:12barrywardelllinuxstb: i just tried brickmania which I think uses bitmaps and it works fine
13:18:21amiconnlinuxstb: That kind of cropping should be handled by the app layer, but specifying LCD_WIDTH and LCD_HEIGHT at maximum
13:18:25theli_uaamiconn, linuxstb, where can i find docs on arm asm which can help in using it with rockbox's arm targets?
13:18:42linuxstbbarrywardell: Yes, it does. So it's odd it doesn't work.
13:18:56linuxstbtheli_ua: Any documentation google gives you about the arm7tdmi
13:19:07theli_ualinuxstb, thnx
13:19:22amiconntheli_ua: You can try to port algorithms from one asm dialect to another, but if the register set / instruction set is too different, it may be impossible
13:20:01amiconnPorting sh1 <-> coldfire <-> arm is not too difficult, but then all 3 platforms are risc(ish) and have a similar register set
13:20:31amiconnx86 is erm, ugly
13:20:55Mikachui have only taken a small course in mips assembler
13:21:15*amiconn has taken no asm courses at all
13:21:38barrywardelltheli_ua: have a look at http://ipodlinux.org/forums/viewtopic.php?t=2311&highlight=arm7 it has some arm docs
13:21:47theli_ualinuxstb, i found a commands description for arm7tdmi in THUMB mode ... does this what i need? (what i THUMB mode? )
13:21:50linuxstbamiconn: I don't understand what you mean by cropping at the app layer.
13:22:25linuxstbtheli_ua: Rockbox doesn't use thumb mode.
13:22:35theli_ua:(
13:22:42amiconnlcd_yuv_blit() takes the source dimensions and the destination rectangle. The app layer should ensure the destination rectangle doesn't exceed the lcd area
13:23:03linuxstbAh, OK. So just the sanity checking of the parameters?
13:23:10amiconnyes
13:23:33linuxstbYes, that makes sense.
13:23:44amiconnChecking every time costs performance (though it's only a small fraction of the whole processing time), and makes the function larger
13:24:19amiconnI'll tackle lcd_yuv_blit() for h300
13:24:53amiconnMan, this video stuff is distracting me from other things :/
13:26:05 Quit merbanan (Read error: 104 (Connection reset by peer))
13:26:29dan_atheli_ua: Have a look at http://www.hitex.co.uk/newsletter0603/arm/Introduction%20to%20the%20ARM%20Architecture_mb.pdf
13:26:57theli_uadoes http://www.gaw.ru/pdf/Atmel/ARM/DDI0029G_7TDMI_R3_trm.pdf alsa helpful? (cause i found this in translation to my second native language)
13:27:15safetydanamiconn, did you get any thoughts on the destroy viewports issue?
13:27:27amiconnNo, not yet
13:27:38amiconnI have some ideas myself, but they're somewhat ugly
13:28:18safetydanIs your concern the need to manually destroy the view ports?
13:28:48amiconnYes that's my main concern, and also how to select the viewports to destroy
13:29:40amiconnThe scrolling code cannot (easily) use the source viewport, it has to crop-and-copy it
13:30:25amiconn..in order to store the current fg/bg, font & style (in the future) etc
13:31:03amiconnMy idea so far was to introduce a viewport ID, and the cloned viewports would all have the same ID as the 'master'
13:31:23*pascals wanders back in
13:31:46amiconnBut then the question remains where this ID is assigned. Doing it in the low-level code would require to initialise and deinitialise a viewport
13:32:02 Join merbanan [0] (i=banan@farmer.campus.luth.se)
13:32:34linuxstbamiconn: Are you planning to use C or assembler for the H300 lcd_yuv_blit()?
13:32:37pascalsCan rockbox do a similar faster-and-faster scrolling as apple firmware does?
13:32:52pascalsI could not find anything about it in the manual
13:33:05linuxstbpascals: Not yet. There's a patch on the patch tracker which tries, but I'm not sure how well it works.
13:33:13pascalsBummer.
13:33:32linuxstbpascals: For an idea of where the ipod port is, check the IpodStatus wiki page - including the long to-do list at the end.
13:33:34amiconnHmm, in fact the init could be done automatically when the app layer uses scrolling text in that viewport. The deinit would be necessary to call only in this case, and would destroy all scroll viewports with the same ID as assigned to the master
13:33:35pascalsIt sucks trying to scroll through 2500+ artists
13:34:12*pascals wishes he had some time to contribute to just fixing some things, not just ask about them
13:34:27amiconnThe reason for this cloning is multiple scrolling lines in one viewport, with potentially different attributes
13:35:23amiconnI think I have to introduce a temporary compatibility layer, until everything is ported to use viewports
13:35:53amiconnAll line oriented text output will go away
13:36:26linuxstbI was expecting viewports to contain line oriented text output - the same way Rockbox works now.
13:36:27amiconnThis will still be compatible with the player, by defining one charcell as one 'pixel'
13:37:17amiconnlinuxstb: What would be the advantage? Imho it's better to have just one way of placing text - by pixel coordinates
13:38:11linuxstbIt's just that that's how Rockbox works now... It means the app doesn't have to care about the font size when displaying text.
13:38:36amiconn[13:32:03] <linuxstb> amiconn: Are you planning to use C or assembler for the H300 lcd_yuv_blit()? <== C
13:39:42Zagoramiconn: wouldn't forcing pixel placement add a lot of code in a lot of places, calculating text sizes?
13:41:11amiconnI don't think so. For viewports you'll have to calculate sizes anyway
13:41:48Zagoryes but only for the first coordinate, not every line of text
13:41:53amiconnPerhaps specifying a line might be useful, but specifying a column doesn't make much sense wrt proportional fonts
13:42:12Zagorthat I agree with
13:42:30barrywardelllooking at the bubbles.c problem there only seems to be a bubbles_bubble in mono not in native so it never gets added
13:42:48linuxstbbarrywardell: Which sim are you compiling?
13:42:52amiconnbarrywardell: What target/sim?
13:43:20barrywardelli have tried my h10 target and ipod mini 1g
13:43:28pascalsSo, all the PortalPlayer info is reverse engineered?
13:43:47linuxstbpascals: Yes, almost entirely by the ipodlinux people.
13:43:53pascalsAwesome
13:44:19pascalsHow are the other hardware manufacturers? Are they willing to provide specs, etc?
13:45:12linuxstbSome do, some don't... But the trend seems to be againt making docs public.
13:45:19linuxstb^against
13:45:24pascalsSad, really.
13:45:43pascalsOTOH, they probably make good money off selling them to their customers.
13:46:21Zagorpascals: actually, the docs are generally free to customers
13:46:36pascals... but under NDA, most likely.
13:47:19Zagorin the PortalPlayer case, yes, but some (like TI) don't even use NDAs. they just don't want non-customers to be able to learn about their products for some unfathomable reason.
13:47:20pascalsHas anyone just gone out and asked to buy the docs from portalplayer and such?
13:47:35amiconnpascals: Our other (current) targets use less integrated solutions, and for most of the individual chips (e.g. sh1 cpu, coldfire cpu, various audio codecs/dacs, etc) docs are publicly available
13:47:41Zagorpascals: yes, I have. they sent me a massive NDA to sign.
13:48:00pascalsZagor: how massive?
13:48:07pascalsAnd what would it have cost?
13:48:10Mikachuif someone signs an nda but publicizes the info anyway, do people using that info break any law?
13:48:24pascalsMikachu: depends on the locale.
13:48:37Zagorpascals: as massive as to be completely incomatible with open source. no cost though.
13:48:57*pascals ponders breaking the law.
13:49:17*pascals decides against it, for now.
13:50:12lexFCK
13:50:13pascalsZagor: do you have the NDA in pdf or text?
13:50:16lexmy usb is broken
13:50:17lexreally
13:50:29lexit gives power to the devices, but any os wont recognize the devices
13:50:45Zagorpascals: not here. maybe I have on my home computer, I don't remember.
13:50:48Mikachutried another port? cold reboot?
13:50:57lexMikachu: anything
13:51:01preglowZagor: so you've actually made portalplayer respond to some request?
13:51:14lexwindows and linux both recognizes the ports but any device won't work
13:51:17Zagorpreglow: yes
13:51:24preglowZagor: not bad in itself, apparently
13:51:37lexhttp://88.112.47.172/dmesg.txt and that's my dmesg
13:51:59Zagorpascals: the NDA is not very special though. just basicall "you can't reveal any information in text or code form"
13:52:29Zagor...plus 20 pages of legal fluff
13:52:35preglowcan anyone give me a good reason why they wouldn't want to reveal specs? do they fear cloners or something?
13:52:54dionoea_worklex: syslog messages when plugging/unplugging the device would maybe help a bit more
13:52:57MikachuZagor: any expiration date?
13:53:21dionoea_workZagor: so you can reveal info in speaches ?
13:53:35Mikachutry sending back a signed nda but modify it to allow you to reveal info, and see if they notice
13:53:44Zagordionoea: no, that was not a literal quote. it's broader than that.
13:53:49lexdionoea_work: nothing there about usb
13:54:00lexonly gconfd stuff
13:54:01ZagorMikachu: I don't remember.
13:54:10dionoea_worktail -f /var/log/syslog . Nothing happens when you plug it in/out ?
13:54:40lexdionoea_work: nothing
13:54:57lexdionoea_work: it's not the first time when this kind of mb brokes
13:54:58dionoea_workis the player on ?
13:55:12lexyes it is, and no usb device wont be recognized
13:55:34dionoea_workdoes lsusb list the device ?
13:55:47lexnope
13:56:08safetydanpreglow, patents and licensing is all I can think of
13:56:10dionoea_workyou have usb kernel modules loaded ?
13:56:18lexAug 8 13:11:27 localhost kernel: [17179932.172000] ehci_hcd 0000:00:1d.7: port 3 reset error -110
13:56:21lexAug 8 13:11:27 localhost kernel: [17179932.172000] hub 5-0:1.0: hub_port_status failed (err = -32)
13:56:25lexhmm
13:56:28lexdionoea_work: yeah, but it wont work on windows either
13:56:31dionoea_worksudo rmmod ehci_hcd
13:56:34lexAug 8 13:11:29 localhost kernel: [17179933.916000] hub 5-0:1.0: Cannot enable port 3. Maybe the USB cable is bad?
13:56:36dionoea_workand plug it back on
13:56:56lexnothing :(
13:57:09dionoea_worksudo modprobe uhci_hcd ?
13:57:23lex[17180826.272000] hub 5-0:1.0: USB hub found
13:57:23lex[17180826.272000] hub 5-0:1.0: 8 ports detected
13:57:24dionoea_workor buy a new cable :D
13:57:32lexhmm wtf
13:57:39lexi have only 6 ports of usbs
13:57:47dionoea_workit's the USB chipset i guess
13:58:34 Quit safetydan ("Ex-Chat")
13:58:44 Quit secleinteer (Read error: 60 (Operation timed out))
13:58:55barrywardelli just copied the bubbles_bubble.*.bmp files from mono to native and added them to the SOURCES
13:59:03lexlsusb shows only five usb ports
13:59:05barrywardellit builds now anyway
13:59:12lexand i have 6 physical ports
13:59:37 Join secleinteer [0] (n=secleint@70.230.186.74)
14:00
14:01:18lexdionoea_work: http://88.112.47.172/syslog.txt and look in 13:11 :)
14:01:44lexnice?
14:01:52lexAug 8 14:56:49 localhost kernel: [17180803.984000] usb usb5: USB disconnect, address 1
14:01:55lexwow!
14:02:40dionoea_workyou should find someone else with a usb mass storage device and try it out
14:02:52lexi have one
14:03:01lexi'm trying with my bluetooth-dongle now
14:03:08lexbut it doesn't recognize it either
14:03:33Mikachutry modprobe usb_storage
14:04:46lexno, i'm getting to understand that the usb is really broken now
14:05:00 Quit PaulJam (".")
14:13:50 Quit Febs (Read error: 145 (Connection timed out))
14:16:36linuxstbMikachu: Could you send me one of your troublesome mencoder files if you still have it?
14:17:41linuxstbMikachu: Thanks.
14:18:21linuxstbIt decodes too fast in mpeg2dec - I can't tell if it is displaying it correctly...
14:18:36linuxstb"896 frames decoded in 0.24 seconds (3733.33 fps)"
14:18:40Mikachuheh
14:18:52Mikachui see, it must sync to display refresh here, i get 85 fps
14:19:04Mikachuit displays correctly here though
14:19:09Mikachubut not on the ipod
14:19:25Mikachuand mplayer test.m2v works, both with ffmpeg and -vc mpeg12
14:19:30Mikachuand i think the latter is the same libmpeg2
14:19:36linuxstbOdd. "file test.m2v" just tells me "data".
14:19:44Mikachuhm
14:20:38linuxstbWas this mpeg-1 or mpeg-2?
14:20:46Mikachumpeg2
14:20:56Mikachubut the method for encoding it was questionable
14:21:08Mikachu-oac copy -ovc lavc -lavcopts vcodec=mpeg2video:vbitrate=300 -o test.avi -vf scale=176:128 -fafmttag 0x55 -ofps 30000/1001 -endpos 30
14:21:13Mikachumplayer -dumpvideo -dumpfile test.m2v test.avi
14:21:39dionoea_workyou get artifacts while playing the file on rockbox ?
14:21:40linuxstbMaybe the dumpvideo option didn't dump it correctly...
14:21:48linuxstbI haven't tested.
14:21:57Mikachuwell, i can encode one with avidemux instead
14:22:02linuxstbCopying it now...
14:23:06dionoea_workMikachu: you get artifacts while playing the file on rockbox ?
14:23:07linuxstbIt's complete nonsense on my 5g - the colours are all wrong.
14:23:20linuxstbIt's just blocks...
14:23:34dionoea_worki had the same issue with a file i encoded with VLC
14:23:41dionoea_workchroma problems
14:23:49Mikachuit looks almost right here
14:24:02dionoea_worksome blocks were right, but some were completely off
14:25:19linuxstbTurning on "-v" in mpeg2dec slows it down - and it looks OK. So it does seem to be a bug in mpegplayer. I'll investigate later.
14:25:50Mikachuthis one has correct `file` output at least
14:26:26linuxstbWas that with avifile?
14:26:32Mikachuavidemux2
14:26:37Mikachuusing lavc for encoding mpeg2
14:26:45Mikachuie ffmpeg
14:26:59linuxstbYes, I meant avidemux2..
14:27:02dionoea_workcan't ffmpeg directly output raw es ?
14:27:24Mikachumaybe it can, but i don't have the ffmpeg client installed
14:29:49 Quit sven_ (Read error: 104 (Connection reset by peer))
14:31:44dionoea_workit could be a difference between http://en.wikipedia.org/wiki/Packetized_Elementary_Stream and http://en.wikipedia.org/wiki/Elementary_stream
14:32:19barrywardelli think i found the problem with compiling bubbles.c on mac
14:33:01barrywardellthe patch sets bmp2rb_mono to be empty so mono bitmaps aren't build
14:33:12 Join sven_ [0] (n=sven@port-212-202-193-252.dynamic.qsc.de)
14:33:15barrywardellgetting rid of that line helps
14:33:46*Bagder has handed over the sansa devboard to Linus to have him try the ARM JTAG wiggler on it
14:33:49barrywardellbut then the build fails when building apps/bitmaps/mono because there are no bitmaps in that directory
14:34:08*preglow rubs hands
14:34:17barrywardelli copied a random bitmap there and everything builds fine
14:35:11barrywardellany suggestions how to get around this problem without adding a bitmap to the apps/bitmaps/mono directory?
14:35:50linuxstbBagder: You got there before he could attach it to his 5g? :)
14:36:38Bagderyou already know how to use your LCD ;-)
14:36:56amiconnbarrywardell: Ah, hmm, now I remember that the OSX library tool has problems with empty libs
14:37:23linuxstbBagder: Barely...
14:37:28Bagderhehe
14:37:39barrywardellamiconn: any thoughts about a good way to get around the problem?
14:37:41amiconnWe should probably have a dummy file that can be compiled and placed in libs for OSX
14:37:52Bagderis there JTAG connections available on the 5g board?
14:38:03amiconnOr we can just require gnu binutils
14:38:23linuxstbBagder: We don't know. There is on the 4g ipods, but the 5g is physically smaller...
14:38:52Bagder"Hello dear Mr Apple, please send us an ipod dev board so that we can jtag it nicely"
14:39:04linuxstbSomeone in IPL has just recently succeeded in attaching a JTAG thing to a 4g.
14:39:09Bagdernice
14:40:27barrywardellamiconn: i think a dummy file would be easier
14:40:29amiconnlinuxstb: Hmm, your lcd_yuv_blit ignores the chroma 'phase' in horizontal direction. So chroma might be off by 1 pixel if the video is wider than the screen
14:40:50amiconnbarrywardell: That still doesn't solve the -Map problem
14:40:57 Join Jillorious [0] (n=50e875a2@labb.contactor.se)
14:40:58*amiconn thinks OSX is a strange OS
14:41:05*Bagder agrees
14:41:43 Quit Jillorious (Client Quit)
14:41:50barrywardellamiconn: I'll try to use -M and redirect the output to a file
14:42:00linuxstbamiconn: I shouldn't do these things at 2am....
14:42:07 Join insanity381 [0] (n=540d4a30@labb.contactor.se)
14:42:29barrywardellwhen OSX "just works" it's great
14:42:39barrywardellbut when it doesn't, it's very frustrating
14:42:52linuxstbFor me, it's close enough to Linux to be usable, but different enough to be frustrating...
14:43:15linuxstbALT-TAB doesn't even work properly...
14:43:16amiconnlinuxstb: It won't be off as long as (video_width - LCD_WIDTH) % 4 == 0
14:44:04amiconnH,, lcd_yuv_blit() doesn't help much on h300 :(
14:44:10amiconn*Hmm,
14:44:41barrywardellgotta go. i'll submit a new patch later
14:44:43amiconn(and there's a minor bug)
14:45:01linuxstbWhat fps gain are you seeing (if any)?
14:45:32amiconnAbout 1 fps...
14:45:48linuxstbIs the lcd_update() already quite optimised?
14:45:48 Part insanity381
14:46:30amiconnyes
14:46:34 Quit barrywardell ()
14:46:40amiconnIt uses burst reading
14:46:47linuxstbIs that a full-screen video, or the smaller 176x128?
14:47:00amiconn176x128
14:47:14linuxstbSo maybe it will help more with a full-size video.
14:49:11amiconnPossibly. I'd need such a video for testing...
14:49:23markunlinuxstb: did you try it yet on a ipod video?
14:49:37preglowMikachu: yo, i'd like to test some encoding of my own. do you have some pre-canned command lines lying around for me?
14:49:52Mikachuhttp://mikachu.ath.cx/rockbox/mpeg2encoding.txt
14:49:57preglowcheers
14:50:47 Quit lex (Read error: 110 (Connection timed out))
14:50:52preglowany idea what package mpeg2enc/dec might be in?
14:50:55linuxstbmarkun: Yes. I've only tried a 320x176 file (16:9 aspect ratio), and that is playing at around 12.5fps
14:50:57Mikachuit's in mjpegtools here
14:50:59preglowa search doesn't yield too much
14:51:04Mikachufor enc
14:51:10Mikachumpeg2dec is in libmpeg2
14:51:15Mikachuand y4mscale is in its own package
14:51:17linuxstbI'm not sure if it's in Debian though - patent issues...
14:51:30Mikachuy4mscaler rather
14:51:43Mikachuyou don't need that if you use mplayer for decoding though
14:51:48Mikachuonly mjpegtools
14:52:22*pascals signs off with a thank-you to the rockbox developers
14:52:36preglowi'm on ubuntu :/
14:53:32daurn|laptopi'm not!
14:54:09dionoea_workpreglow: debian has a mpeg2dec package
14:54:20dionoea_workyou might need to grab that in the universe repository in ubuntu
14:54:35preglowyeah, i see, but i still need mpeg2enc
14:54:47Mikachuin a pinch, you could just compile it?
14:54:50preglowsure
14:55:15Mikachusome example program didn't compile for me, it wanted linuxthreads/semaphore.h, but it seems to not be needed for anything
14:55:15 Quit pascals ()
14:55:17dionoea_workjust grab it of svn://svn.videolan.org/libmpeg2/trunk
14:55:46preglowweird
14:55:49 Join perpleXa [0] (n=perplexa@unaffiliated/perplexa)
14:55:49preglowi've got libmpeg2
14:55:53Mikachui didn't know vlc owns libmpeg2
14:55:53preglowstill no mpeg2enc
14:56:04Mikachuit is like debian to split packages confusingly
14:56:05linuxstbdionoea: Is that the same libmpeg2? My version is from http://libmpeg2.sourceforge.net
14:56:08dionoea_workpreglow: it's in the source code
14:56:13preglowdionoea_work: roit
14:56:16dionoea_workthe sourceforge CVS is down
14:56:23dionoea_workso sam decided to move it to videolan.org
14:56:25linuxstbThey have nightly tarballs though.
14:56:34linuxstbSo those tarballs might be out of date?
14:56:42dionoea_workbut it's not very active these days i think
14:56:45Bagdersourceforge cvs is usually more down than up
14:56:53Mikachudon't nightly tarballs usually have dates on them?
14:57:05linuxstbNot the one I downloaded...
14:57:13linuxstbIt just said "0.4.1-cvs"
14:57:27linuxstbThe last official release was 0.4.0b
15:00
15:01:25markunlinuxstb: can't wait to try video playback on the gigabeat..
15:01:43linuxstbmarkun: You're spoilt with your 300MHz ARM...
15:02:12linuxstbI'm guessing you'll need to rotate the LCD though.
15:02:21markunor rotate the movies :)
15:02:37linuxstbIs that a standard encoder feature?
15:02:38linuxstb:)
15:02:48markunmencoder can do it
15:02:53 Quit Zagor ("Client exiting")
15:02:55linuxstbOf course it can...
15:04:04Mikachurotate, flip, mirror, perspective correction
15:04:20Mikachuif you want to hold your screen at a slight angle? :)
15:04:27markun:)
15:07:37***Saving seen data "./dancer.seen"
15:08:01amiconnlinuxstb: Btw, as-is the greyscale ipods probably wouldn't even manage realtime video with rvf + audio
15:08:14 Join Poka64 [0] (i=peter@hd5e241ae.gavlegardarna.gavle.to)
15:08:19amiconnFeeding the frames to the lcd takes a lot of cpu on ipod.
15:08:30amiconnNow if the grayscale lib could use the cop...
15:08:49*dan_a wakes up
15:09:39amiconnIt would mean running isrs on the cop too
15:10:41amiconnAre the vectors shared?
15:10:52preglowyes, i think so
15:11:01amiconnhmm :/
15:11:09dan_aVectors are shared, but I'm having a horrible time trying to get timer interrupts working
15:11:16preglownot exactly a showstopper
15:11:24preglowjust check which core you're at and branch to another handler
15:11:28Mikachudoes the lcd controller require that much cpu or is it just the only known way so far?
15:11:30dan_a(by working I mean "not freezing rockbox")
15:13:21 Join Criamos [0] (n=Criamos@p54933702.dip0.t-ipconnect.de)
15:14:57amiconnMikachu: The transfer takes rather long because the cpu has to wait for a 'ready' signal before sending more data. If we knew how to program the PP's interface to the lcd controller, it could go a lot faster
15:15:55amiconnI don't believe the timing needs to be really as slow as it is now. The H1x0 lcd controller manages at least 10x the data rate
15:16:22Mikachuyeah i remember looking at the code once
15:16:42amiconns/at least/about/
15:17:37Mikachuwould be nice if the sansa could reveal something there
15:21:19 Quit jaczehack (Read error: 110 (Connection timed out))
15:22:52 Quit dionoea_work ("CGI:IRC (EOF)")
15:25:02theli_ualinuxstb, i've played today with zxbox code a bit ... now i can run chuckie the egg with 16 fps with cpu unboosted and around 41 fps with boosted
15:28:24 Quit tommeyer (Read error: 110 (Connection timed out))
15:28:25linuxstbtheli_ua: What player is that on?
15:28:50theli_uamini
15:29:00theli_ua2g
15:30:29linuxstbSo is 25fps realtime?
15:31:05theli_uayes, i think so
15:31:12linuxstbYou could try the greyscale lib...
15:31:22theli_uaso with boosted cpu i get 41 fps without frameskip
15:31:39theli_uait doesn't look nice yet on ipods to my taste :)
15:31:51theli_uabut that's a nice idea
15:32:02theli_uai'll try it now
15:32:33linuxstbamiconn: Did you try that 224x176 test file on your H300?
15:33:01amiconnlinuxstb: 224x176 test video: 7.33fps with cvs, 7.78fps with lcd_yuv_blit()
15:33:19 Join lex [0] (i=lex@193.65.221.148)
15:33:23amiconn(average across the first 8 fps values)
15:33:46linuxstbThe last fps displayed should be the most accurate.
15:33:48amiconn6% speedup
15:34:07amiconnOh?
15:34:30amiconnThat's strange, because it means decode speed increaes even more than the fps would suggest
15:34:33linuxstbI think full-screen on the Nano went from about 18fps to about 24fps with the blit function - but that's due to those waits in the lcd update.
15:35:34linuxstbThe fps calculation is in video_out_rockbox.c - please tell me if it's doing something stupid, or if you can think of a clever way to improve the accuracy.
15:35:56amiconnThe %125 is wasteful, because it's not a power of 2
15:36:05linuxstbYes, I though of that this morning...
15:36:27amiconnWhy not display a stats screen at the end instead of showing the fps during playback?
15:36:35linuxstbYes, it should do.
15:36:49amiconnIf the top left corner is covered by the video, it's quite hard to catch the fps values
15:37:05Mikachuit's not supposed to be covered i think :)
15:37:19amiconnWith full screen video it is
15:37:42linuxstbMy tests have mostly been with 16:9 ratio files, so the fps stays visible.
15:38:09 Quit Criamos ("( www.nnscript.de :: NoNameScript 4.02 :: www.XLhost.de )")
15:41:23linuxstbdan_a: Have you tried without enabling the cache - to make sure that isn't causing you problems?
15:41:59dan_alinuxstb: Yes, the cache is off on both cores, and I turned off frequency scaling too (that caused extra problems)
15:42:33dan_aI think the trouble is with my logic of what I need to do at a timer interrupt
15:43:09dan_a...and with what variables should be local to the core, and what should be shared between them
15:43:53linuxstbThinking about the video player, all it needs from the COP is to be able to run a function on it. Some simple communication can happen via IRAM.
15:44:14Mikachubefore the yuv_blit the fps was displayed over the video at least
15:44:31Bagderisn't that mainly what the COP should do? wait for inputs, run a function, deliver result, wait for new input
15:44:55dan_aI think so.
15:45:14linuxstbMikachu: That's because the video was drawn into Rockbox's lcd framebuffer, and then the fps drawn on top, then the framebuffer was written to the LCD hardware. The yuv_blit just sends the video directly to the LCD hardware, bypassing Rockbox's framebuffer.
15:45:51amiconnlinuxstb: I've just committed H300 lcd_yuv_blit()
15:46:48linuxstbNice.. I'll commit my 5g later, then we'll just need a version for the X5. I'm disappointed with the result though.
15:47:05*amiconn is still waiting for his X5 to arrive :(
15:49:15Bg3rany cyrillic amarok user here ?
15:49:29dan_aDo you think we need some kind of mailbox API for passing data between the processors, or will the queueing code be enough?
15:50:11Bagderdan_a: it'll show as soon as we start making us of it
15:50:31linuxstbDon't build the bridge until you've reached the river...
15:50:44*dan_a has dabbled his toes
15:51:03 Quit sven_ ("Leaving")
15:51:09 Quit gtkspert ("leaving")
15:51:20 Quit perpleXa ("Leaving")
15:52:05dan_aI've already tried making the codec thread run on the COP, but can't make it work properly
15:52:20 Join gtkspert [0] (n=gtkspert@124-168-73-220.dyn.iinet.net.au)
15:52:22amiconnhuh?
15:52:40*amiconn is confused about the commit history display & build table
15:53:17 Quit chendo (Remote closed the connection)
15:53:21amiconncvs told me that the up-to-date check failed on 2 files, so I wouldn't have expected it to execute the commit for the other 3...
15:53:41 Join chendo [0] (n=chendo@203-206-49-75.dyn.iinet.net.au)
15:54:05Bagderbtw we have a plan for our step towards svn
15:54:17JdGordonyou are going to svn?
15:54:25amiconnBagder: Please compare http://www.rockbox.org/cvsmod/chlog-2006-08-08%2013:45:55.html with the frontpage
15:54:37linuxstbBagder: Have you definitely decided on svn?
15:54:41amiconnWhy are plugin.[ch] only displayed on the former?
15:54:59 Join dionoea_work [0] (n=3e179bf9@poy.chewa.net)
15:55:46*linuxstb recalls Zagor mentioning git
15:55:50Bagderlinuxstb: I think so, since we don't really feel like doing a lot of testing and comparing and svn is a step forward
15:56:22Bagderamiconn: haha, check now ;-)
15:56:44Bagderit fixed itself
15:57:06Bagderwhich of course is because the scripts ran at different times
15:57:18amiconnhuh?
15:57:44amiconnThese scripts are odd...
15:57:50Bagdernot really
15:57:55BagderCVS is the one that's odd
15:58:01Bagderthe scripts work hard to "fix" that
15:58:26linuxstbBagder: So what is the plan for svn?
15:59:09Bagderlinuxstb: we cvs2svn to a new repo on a new host and setup all scripts to work on that, then we stop cvs, do the conversion again for real and use the new place for real
15:59:19Bagderand we never look back again
15:59:29*linuxstb prays
15:59:33Bagderoh, and 3) profit!
15:59:57JdGordondont forget the dragons...
16:00
16:00:05JdGordonhow long will 18gb take to copy to usb?
16:00:15Bagderand the svn repo will be hosted by the download.rockbox.org site
16:00:36linuxstbDo you know how people can deal with local trees with uncommitted changes?
16:00:44linuxstb(apart from committing...)
16:01:06linuxstbI suppose make a diff, and apply the diff against a new svn checkout...
16:01:35Bagderyes
16:01:45amiconnCan't you just tell svn to only include certain files in a commit, like you can with cvs?
16:02:12dionoea_workyou'll still have to merge changes done while it was using CVS
16:02:23dionoea_work*local changes
16:02:33amiconnAh, hmm
16:02:36linuxstbBagder: It would be useful to keep the cvs server alive (for creating such diff), but just disabling write priveleges (if possible).
16:02:42Bagdertrue
16:02:45amiconnlinuxstb: You mean one-time, when doing the conversion?
16:02:59*amiconn probably misunderstood sth
16:03:35linuxstbYes - I've got a few local cvs trees with changes I don't want to commit yet. So I'll need to transfer them to local svn trees.
16:03:51amiconnI have some local changes which should never be committed
16:04:08dionoea_work(and then you can install trac to do project management, wiki and svn browsing)
16:04:19JdGordonto format the usb disk as fat32 do i use mkfs.msdos or .vfat?
16:04:24amiconn(.e.g. disk poweroff for the player - it works fine on mine, but not on all)
16:04:37linuxstbJdGordon: I think they're just aliases for each other. But I use vfat.
16:04:44amiconnIt's just a #define in config-player.h
16:08:22 Join KlrSpz [0] (n=klrSpz@69.15.248.2)
16:12:47*amiconn hopes linuxstb will do what his commit msg suggests
16:12:57*t0mas is now playing: Joe Satriani - Surfing With the Alien
16:13:00t0masah man, wonderful song
16:13:39t0masI never understand how he gets this sound from a guitar... but still... it's great :D
16:14:22linuxstbamiconn: What did I suggest?
16:14:35amiconnAh, that was ambiguous
16:14:43amiconnI meant my suggestion
16:14:46linuxstb:)
16:14:49 Quit BHSPitLappy (Connection timed out)
16:15:07linuxstbAh, the bounds check? Yes, I'll do that.
16:15:11amiconnyes
16:18:06 Join mirak [0] (n=mirak@ip-7.net-82-216-143.rev.numericable.fr)
16:18:26miraksalut
16:18:33linuxstbHi.
16:18:58linuxstbI've been meaning to ask how your xvid decoder compared to your libmpeg2 decoder. Was libmpeg2 better?
16:21:28linuxstbI seem to recall that the .rock file was huge - mpegplayer.rock is only about 75KB for the ipod.
16:21:56JdGordonhow many blocks are in a 20gb fat32 disk? i told it to do a bad block check before formatting but im bored and its up to 8000000 already
16:21:59amiconn~65KB for H300
16:22:35linuxstbAnd aac.codec is about 250KB...
16:22:47 Join elinenbe [0] (n=d1c4c007@labb.contactor.se)
16:22:49linuxstbEven vorbis is bigger than mpegplayer...
16:23:20amiconnStrange... aac.codec for coldfire is larger than for ipod
16:23:29linuxstbAnd slower...
16:23:39amiconnyup :(
16:23:58amiconnNot that I'd use aac, but a half-finished codec isn't good...
16:24:39linuxstbIt's usable on the ipods though, so I wouldn't want it removed.
16:24:53linuxstbBut I agree it's not nice to see it more or less abandoned.
16:25:26elinenbelinuxstb: nice work on the mpeg... I'm liking it. A LOT!
16:25:26amiconnEveryone is waiting for ffmpeg's aac implementation I guess...
16:25:46elinenbeamiconn: what's the status with ffmpeg's aac?
16:25:55amiconndunno...
16:25:56linuxstbWhich if it turns out to be floating-point only may not be much help...
16:26:32linuxstbelinenbe: Someone is working on it this summer, sponsered by Google's Summer of Code. But I'm not sure if it will be floating-point only (i.e. no use for Rockbox), or if there will also be a fixed-point version.
16:26:53linuxstbThe first implementation uses floating point.
16:26:58markunelinenbe: there is some code: http://svn.mplayerhq.hu/aac/
16:28:23elinenbeah... thanks.
16:28:41 Quit secleinteer (Read error: 60 (Operation timed out))
16:29:39 Join secleinteer [0] (n=secleint@adsl-70-237-211-95.dsl.stlsmo.sbcglobal.net)
16:31:09miraklinuxstb: yes the problem was the size of the file. I needed to increase plugin ram size. I tried to remove many functions, I am sure more could be removed. The performances where just a bit better on mpeg2. something like 2.4 fps against 2.8 (at 45mhz cpu)
16:34:06miraklinuxstb: with xvid there was a lot of unused conversion fonctions. However I think that xvidcore was more easy to optimise for iram than libmpeg2. xvidcore used little buffers to do the idct, on libmpeg2 it's less clear, it's more like if it works directly on a buffer the size of the image
16:34:55miraklinuxstb: also I remember than on xvid the functions used to parse the stream where badly designed
16:35:47miraklinuxstb: you would do a lot better than me for xvid I think
16:36:35mirakone thing also is to know if we want to keep most of the functionality of the codecs, like scaling, different colors converstions etcetera
16:37:18linuxstbNo, I don't think we need scaling or colour conversions in the codecs. Assuming xvid uses YUV, then it should hopefully be able to use the lcd_yuv_blit() functions we're starting to write.
16:37:45linuxstbSo we just need the codec to give us a yuv bitmap.
16:38:01linuxstb(and the user should re-encode to the size of the LCD)
16:38:22elinenbeI don't think it is unreasonable to have the user convert to the size of the LCD
16:38:54dionoea_workxvid is mpeg4 which uses YUV like mpeg2
16:39:17mirakfrom your experience is it faster to copy a a memory block on wich an idct is performed to iram and work on it, then swap it back again, to slow ram, ?
16:39:25 Join chendo_ [0] (n=chendo@203-206-76-190.dyn.iinet.net.au)
16:39:55mirakor work directly on slow ram. I am wondering at wich point it's better to do a kind of swapping to iram
16:39:58linuxstbmirak: I've no experience in that.
16:40:17linuxstbI think we would just need to test.
16:40:34mirakelinenbe: actually there is not much choice, the devices aren't powerfull enough to scale
16:40:54linuxstbAnd also it would mean decoding much more data than is needed.
16:41:41mirakconverting a move takes less than 30 minutes. That's a kind of thing that you can do in advance before a train trip. I did it some times
16:41:54miraka movie
16:42:43miraklinuxstb: anyway that's really fine to have the video plugins in cvs. I am sure it will attract some hackers
16:44:40linuxstbmirak: Hopefully.
16:47:20dionoea_worki don't know if you could do something like only update parts of the screen which don't change between frames... (maybe that doesn't happen all that often)
16:47:51 Join Paul_The_Nerd [0] (n=Paul_The@cpe-70-112-165-230.austin.res.rr.com)
16:48:15dionoea_workbtw, what's the status on JdGordon's keyboard actions patch ?
16:49:04Paul_The_NerdAnyone in here tried the m2v player plugin? Out of curiosity, does it regularly look significantly worse than the same file played on a PC, or am I dealing with an "It doesn't like the program I used to encode" issue?
16:49:18dionoea_worksecond issue
16:49:23Paul_The_NerdGotcha
16:49:27linuxstbffmpeg?
16:49:38dionoea_workyou have to use linuxstb's magical command line to make it work nice :)
16:49:56Paul_The_NerdI grabbed some point and click utility, since I'm in windows, but it seems like it probably used ffmpeg.
16:50:07excitatoryi seem to get significantly faster transfer rates when dumping files onto my nano when it is fully charged. would that make sense, or is it just my perception?
16:50:26linuxstbThe good thing is that mpeg2dec plays those files - and mpegplayer is basically a port of mpeg2dec.
16:50:56dionoea_worki think it's a es / pes issue
16:51:03Paul_The_NerdShould the framerate be displayed on top of, or underneath, the playback?
16:51:33dionoea_workupper left
16:51:34linuxstbIt _should_ be displayed on top... But it's get overwritten if you're playing a full-screen video.
16:51:48Paul_The_NerdGotcha
16:52:01Paul_The_NerdYeah, it was getting overwritten with mine, and I was curious if that was expected too.
16:52:11linuxstbBut I'll probably do what amiconn suggested, and replace it with a stats screen at the end of playback.
16:55:16Mikachucan't you do the fps in yuv space? :)
16:55:19 Quit chendo (Read error: 110 (Connection timed out))
16:55:22amiconnA stop button would also be useful...
16:56:24 Join qwx [0] (n=qwm@h229n1fls34o1010.telia.com)
16:56:51 Nick daurn|laptop is now known as daurn|afk (n=quae@unaffiliated/daurnimator)
16:56:59markunDoes someone have a screenshot of rockbox playing a video? (so I can tease the mygigabeat people)
16:57:07linuxstbMikachu: Patches welcome :)
16:57:24linuxstbCheck the logs - I think someone posted one this morning.
16:58:05daurn|afka video, of a video, of a video, of a video
16:58:07daurn|afk;)
16:58:19dionoea_workMikachu: you'd basically need to write a font rendering function which changes Y to 0 or 255 where you want text. But i'm not sure that you'd want to overlay stuff on the video atm ... takes CPU power :)
16:59:26Paul_The_Nerdmarkun: Take a screendump of Rockbox displaying a 176x128 jpeg. :-P
16:59:35markunlinuxstb: looks terrible, but thanks :)
17:00
17:00:00markunPaul_The_Nerd: doesn't look very good on my H120 :)
17:00:21Paul_The_NerdHahaha
17:00:38linuxstbmarkun: It should work in the sim...
17:01:11dionoea_workdid someone already test it on greyscale LCDs ?
17:01:11amiconnIt *does* work in the sim
17:01:34amiconnGor 110fps and higher
17:01:37amiconn*Got
17:03:30amiconndionoea: Nope, and it seems rvf would be more suitable for these
17:03:52amiconnNow if I could figure out how to use libmad and pcm playback...
17:04:02*amiconn wonders where Wett disappeared
17:04:18dionoea_workwell you could just output the y component with the 32 levels of gray lib ... it might even look ok
17:04:45amiconnHe said he already had rvf working on H1x0, including sound, but then disappeared and published no code... :(
17:05:02dionoea_work:/
17:05:10Bagderits not the first time that happens
17:05:22Bagderremember the guy with the archos emulator?
17:05:28amiconnno?
17:05:39Bagder cpu, ata and lcd emulation
17:05:57Bagderthen the guy vanished
17:06:10amiconndionoea: It will be rather grainy because the redraw interferes with the temporal dithering in the grayscale lib, as they're not synchronised. And while being more cpu intensive, it wouldn't save much on data rate
17:06:40 Quit qwm (Connection timed out)
17:06:45dionoea_workhum ... indeed
17:06:50amiconnrvf on oour greyscale targets (H1x0, ipod g3, g4 and minis) will be around 1.4..1.6Mbps (for the video part), so it should be ok
17:07:01Bagderand I won't be surprised if we never see any code for the currently claimed wma codecs
17:07:05amiconn(That's about the same data rate as CD quality audio)
17:07:41***Saving seen data "./dancer.seen"
17:07:45 Quit theli_ua (Remote closed the connection)
17:07:52ShadowdogMUHey, album art (from iTunes), is stored in the IDv3 info right?
17:07:58amiconnIf we stick with mp2/mp3 for the audio track, only a few modifications are necessary in the rvf toolchain
17:08:26amiconnMostly the different resolution, and the different pixel packing for ipod
17:08:32mirakdionoea_work: "parts of the screen which don't change between frames" in fact that's what mpeg do itself. detecting what doesn't change is what a step in the compression scheme and it takes a lot of ressources to do that
17:08:48amiconnNow if I knew how to deal with libmad and the pcm playback api :/
17:08:52dionoea_workmirak: i know ... but you still redraw the full screen at every frame
17:09:09dionoea_workso i was wondering wether you could still redraw only the part changed in P and B frames
17:09:56ShadowdogMUIs it possible for me to get the album art from a Mp3, I accidently delete the files from iTunes, but I had my backup RAR with all of them in it
17:11:56lexShadowdogMU: there's no album arts on mp3s
17:12:03lexit's an external image
17:12:43ShadowdogMUOk, well the RAR has the /jpgs, I see them, but when I extract them, they aren't there
17:12:57amiconnStrange...
17:13:09*amiconn is now very tempted to work on rvf playback
17:13:13amiconnH1x0 first
17:14:20Mikachuit is possible the art is in the mp3 file, but it's a bit silly
17:14:35*linuxstb wonders what happened to Rockbox being an audio player...
17:14:45dionoea_workShadowdogMU: maybe they're hidden files
17:14:59ShadowdogMUAlready checked
17:15:13ShadowdogMUIf I re-add them through the extracted folder... it has the album art
17:16:27 Join rconan [0] (n=richard@82.14.27.61)
17:16:39linuxstbamiconn: How does the Archos video player buffer the data? Does it demux it into separate audio/video buffers as it reads from disk?
17:16:48amiconnnope
17:17:09amiconnThat would be difficult, as the audio track might be vbr
17:17:29amiconnBut the audio blocks always contain complete frames iirc
17:20:27amiconngah
17:20:38amiconnrvf_mux is C++
17:20:58Mikachualso, is the vkeyboard still missing ; and :?
17:23:20 Join lightyear [0] (n=lightyea@p548778E0.dip.t-dialin.net)
17:25:30 Join BigMac [0] (n=dummeska@64-252-111-214.adsl.snet.net)
17:33:20 Quit einhirn ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
17:40:13amiconnlinuxstb: I just checked; the audio blocks do _not_ always contain complete frames.
17:40:31amiconnBut there is a pointer indicating the last frame boundary within the audio block
17:42:34amiconn(and another one indicating the first frame boundary)
17:42:51linuxstbI think MPEG program streams do that...
17:43:27amiconnIs it possible to give libmad buffers with incomplete frames, and the rest of the frame in the next buffer?
17:44:05dionoea_workshouldn't the demux make sure that the decoders get full frames only ?
17:44:38amiconnThis would be an extra step... I'm talking about .rvf here
17:47:04 Join pixelma [0] (i=pixelma@82.193.235.34)
17:49:26 Join Bger [0] (n=bager@rockbox/developer/Bger)
17:57:24amiconnThe 'most recent changes' in the wiki don't work properly
17:57:50Mikachuhttp://www.rockbox.org/twiki/bin/view/Main/WebChanges ?
17:58:33Mikachuoh, the box at the bottom
18:00
18:03:05 Part Paul_The_Nerd
18:10:40 Quit rconan (Client Quit)
18:11:08linuxstbMikachu: Your test files work fine in the sim....
18:12:04linuxstbSo I'm guessing it's a bug in my lcd_yuv_blit...
18:13:10 Join webguest26 [0] (n=54995f50@labb.contactor.se)
18:14:58 Join Jungti1234 [0] (n=jungti12@125.188.3.64)
18:15:09webguest26is there any chance that the mp3 decoder could be looked at, as i get the accassional artifact during playback om my H140, which i dont get with original OS all my ,usic is encoded with LAME@aps, thanks
18:16:19 Quit mirak (Remote closed the connection)
18:16:41 Part webguest26
18:16:50Jungti1234hahaha..
18:16:52Jungti1234hello guys
18:17:00Jungti1234http://www.cdpkorea.com/zboard4/data/freeboard/cdpkorea-1155048907-1.jpg
18:19:13 Quit thegeek (Read error: 104 (Connection reset by peer))
18:19:15Mikachulinuxstb: the weird thing is the video gets good sometimes
18:19:21Mikachuin parts of the image
18:20:53 Join Davide-NYC [0] (n=chatzill@user-12hdtm5.cable.mindspring.com)
18:24:17 Join jaczehack [0] (n=jac@c-d0f772d5.013-1-73746f40.cust.bredbandsbolaget.se)
18:26:28*linuxstb thinks that perhaps webguest26 should have given more info....
18:26:41Jungti1234hm
18:26:52lexhei linuxstb
18:26:55lexor Mikachu, someone :p
18:27:04lexhow can i convert a mpeg-1 on linux with sound?
18:27:25Mikachuto what/from what?
18:27:54lexto mpeg1 from xvid
18:28:03lexi'd like to have sounds on my videos too
18:28:06Mikachuyou are aware the mpegplayer only plays video at this point?
18:28:32lexi read somewhere that it plays mpeg1 with sound and mpeg2 without sound
18:28:49linuxstbNo... Who's been spreading misinformation?
18:28:52Mikachusomewhere?
18:28:55lexi don't remember
18:29:06lexi'll check
18:29:06Mikachutry
18:29:27lexor maybe i readed incorrectly
18:29:31lexIt currently only plays raw MPEG-1 or MPEG-2 video streams (no audio).
18:29:38lexit could be saying that both have no sound :o
18:30:08Mikachuthat is in fact what it is saying
18:30:39lexyeah
18:30:54lexi read it like it says that mpeg-1 has sound but mpeg-2 doesn't
18:33:42linuxstbBut no - it only plays 1) raw MPEG-1 video with no audio; and 2) raw MPEG-2 video with no audio.
18:34:34 Join Paul_The_Nerd [0] (n=Paul_The@cpe-70-112-165-230.austin.res.rr.com)
18:35:38 Join mikearthur [0] (n=mike@82-41-205-190.cable.ubr11.edin.blueyonder.co.uk)
18:40:26 Part Paul_The_Nerd
18:40:30 Quit dan_a (Read error: 110 (Connection timed out))
18:41:44 Quit dionoea_work ("CGI:IRC (EOF)")
18:44:11 Quit elinenbe ("CGI:IRC (EOF)")
18:45:17 Join petur [0] (n=petur@rockbox/developer/petur)
18:46:26 Quit secleinteer (Read error: 60 (Operation timed out))
18:47:22 Join secleinteer [0] (n=secleint@70.230.181.91)
18:51:43 Join thegeek [0] (n=thegeek@s026b.studby.ntnu.no)
18:52:39Davide-NYCpetur, do you know if any of the currently available AGC patches are synced with the CVS?
18:53:12linuxstbMikachu: Seems the problem is elsewhere - running the same LCD update code on my ipod that the SIM uses (i.e. not the new blit function) still gives me a corrupt image... So it seems the actual decoding is failing on the ipod, not the display...
18:53:23Mikachui see
18:54:18Mikachuthat is a bit odd
18:54:24linuxstbBut the files seem perfect in the sim - I've add a sleep(HZ/25) to it, so it displays at 25fps.
18:55:02Mikachutry gcc -O0 maybe?
18:55:34linuxstbWorth a test
18:56:59linuxstbNo, no difference.
18:57:09linuxstbBut much slower...
18:58:58 Join Febs [0] (n=40be240f@labb.contactor.se)
18:59:55Mikachuyou also see the corruption on your ipod, but it works on other targets?
19:00
19:00:06Mikachui guess i won't reach any conclusion from this, i'm just curious
19:01:26linuxstbYes, I get corruption on both my ipods (Color and 5g).
19:02:52Mikachunot boosting doesn't help either
19:04:08linuxstbI think we would have more serious issues if boosting caused things to go wrong...
19:04:20*linuxstb wonders if it's signed/unsigned chars...
19:04:22Mikachuwell, it did happen to me when i increased the freq from 75 to 78
19:04:32Mikachuassuming i didn't accidentally increase it way more
19:04:51Mikachui thought i might as well try it
19:05:31Mikachui see only "char" and "unsigned char" in the code, what is default for arm?
19:05:57linuxstbI forget.... It's whatever everything else isn't :)
19:06:01Mikachuheh
19:06:14linuxstbI think it's signed char on ARM.
19:06:44linuxstbBut no, libmpeg2 is using the int8_t etc types everywhere.
19:07:45***Saving seen data "./dancer.seen"
19:10:42Mikachui don't want to sound stupid, but where in the Makefile is the command for the .o files defined?
19:10:48Mikachuis it inherited only?
19:11:57Mikachuokay, it uses the CFLAGS at the top anyway
19:12:05BagderMikachu: tools/make.inc
19:12:06linuxstbinclude $(TOOLSDIR)/make.inc
19:12:22Bagderuse 'make V=1' to see the command lines
19:12:34Jungti1234bagder
19:14:08Mikachuand the last -O option takes precedence in gcc right?
19:14:37linuxstbI believe so, yes.
19:14:44Mikachubecause the $(GCCOPTS) includes -O
19:16:28 Join barrywardell [0] (n=barrywar@host-194-46-234-58.dsl-ie.utvinternet.net)
19:16:30 Quit secleinteer (Operation timed out)
19:16:43 Join midgey34 [0] (n=TJ@c-71-205-31-207.hsd1.mi.comcast.net)
19:17:39 Join secleinteer [0] (n=secleint@adsl-70-237-142-251.dsl.stlsmo.sbcglobal.net)
19:17:49Davide-NYCpetur?
19:18:01*linuxstb disappears for a while
19:18:13*petur pops in again
19:19:42peturDavide-NYC: my agc patch misses some changes by Mmmm, and Mmmm's patch is a but broken
19:19:51peturs/but/bit
19:20:16Davide-NYCso I should go with the patch from 20060803 correct?
19:20:23Davide-NYC(for now)
19:20:31Mikachuautomatic gain control?
19:20:36Davide-NYCyes
19:20:43peturI'll put up my version, ok?
19:21:07Davide-NYCit sync with CVS?
19:21:23peturyes... diffing now...
19:21:28Davide-NYCawesome, thanks
19:22:19Davide-NYCpetur you have H300 or H120?
19:22:25peturbleh... somethings wrong...
19:22:27peturh340
19:22:46Davide-NYCyou have Joystick remote?
19:22:58peturno
19:23:11peturjust a h100 one that whatboutbob gave me
19:23:53Davide-NYCif I can find one for less than a small fortune I'll send you a joystick remote
19:24:05Davide-NYC:-)
19:27:02peturpatch is on the tracker
19:27:17*petur pops out again
19:32:16 Join Isobutane [0] (n=5224eef1@labb.contactor.se)
19:32:36IsobutaneHey
19:33:04IsobutaneCan anyone spare five minutes for me?
19:33:55Jungti1234why?
19:34:54IsobutaneI'm trying to put the firmware into the pacth thing but I can't find the .hex and just keep opening folders to blank folders
19:34:59IsobutaneStupid I can't find it, I know
19:40:21markunIsobutane: 'patch thing', 'blank folders'... what are you trying to do?
19:42:07Isobutanerockbox an iriver H320
19:42:20markunhow far did you get?
19:42:23IsobutaneI'm using the guide on the site, have downloaded the firmware and patcher
19:42:25Isobutanenot very
19:42:34IsobutaneI'm browsing with the patcher for the .hex, which I can't find
19:43:53dionoeare
19:48:44dionoeais pcmbuf.c the file to look at if i want to get decoded audio right before it gets sound to the audio output hardware on sw codec platforms ?
19:49:29Mikachugets sent?
19:49:37dionoeawell get the raw audio basically
19:50:11dionoeathe last function before hardware takes over
19:50:32Jungti1234bye
19:50:44 Quit Jungti1234 ()
19:52:53 Quit Isobutane ("CGI:IRC (EOF)")
20:00
20:00:08 Join Nibbler [0] (n=sven@port-212-202-193-252.dynamic.qsc.de)
20:00:27 Quit Nibbler (Client Quit)
20:01:00*dionoea tries to find a wiki page explaining the audio decoding process in rockbox
20:01:34 Join qwm [0] (n=qwm@h229n1fls34o1010.telia.com)
20:05:12dionoeahum firmware/pcm_playback.c looks nice too :)
20:05:28 Join mirak [0] (n=mirak@ip-7.net-82-216-143.rev.numericable.fr)
20:07:31*amiconn just found out that gfx on linux seem to be slower than on windows
20:07:32 Join Nibbler [0] (n=sven@port-212-202-193-252.dynamic.qsc.de)
20:07:37 Quit Nibbler (Read error: 104 (Connection reset by peer))
20:07:55Bagderamiconn: that's common indeed
20:08:21amiconnAt least sdl... With mpegplayer I get 110fps+ on windows (1.5GHz Pentium M) but only 45fps on linux (Athlon64 3800+)
20:08:48 Join RoC_MM [0] (i=dragon@c-24-129-89-52.hsd1.fl.comcast.net)
20:09:52Mikachui don't think sdl uses anything fancy like the render extension or cairo...
20:09:57Mikachubut on windows it probably uses directdraw
20:12:54 Quit qwx (Read error: 110 (Connection timed out))
20:13:39amiconnlinuxstb: I probably found the bug...
20:15:21 Join Nibbler [0] (n=sven@port-212-202-193-252.dynamic.qsc.de)
20:17:00secleinteerhey are any ipod photo devs around?
20:17:31BigMacyah right here
20:17:46secleinteeri found a problem yesterday
20:17:55secleinteerif i'm charging my ipod while playing music
20:18:06secleinteerand the battery level reaches 100%, rockbox just freezes
20:18:15BigMaci was just joking
20:18:19secleinteeroh lol
20:18:21BigMacbut you can fill out a bug report
20:18:22secleinteergoddamit
20:18:25secleinteerwhere?
20:18:29BigMacill link ya
20:18:32secleinteernvm got it
20:18:42BigMack
20:19:05secleinteerwait u have to register to report it?
20:19:14secleinteerscrew that
20:19:46BigMacwhy?
20:19:48markunsecleinteer: the devs are reading the irc logs
20:19:53secleinteerk
20:20:00Mikachuamiconn: and?
20:20:03BigMachow bout you write down what happens
20:20:07BigMacand send it to me
20:20:14BigMacand ill post it for ya
20:20:20secleinteerread above
20:20:26secleinteer if i'm charging my ipod while playing music
20:20:32secleinteerand the battery level reaches 100%, rockbox just freezes
20:20:48markunsecleinteer: are you using any patches?
20:21:05secleinteerno
20:21:13secleinteeri don't think i've added anything
20:21:19secleinteerexcept maybe a font or wps
20:21:45secleinteeri'm using a wps available in the default package, as well
20:22:02 Quit _Lucretia_ (Read error: 113 (No route to host))
20:22:08markunok, just wanted to make sure
20:22:17mirakamiconn: it depends if sdl uses overlay or not
20:22:41markunsecleinteer: sometimes people are having problems with a patched build which doesn't exist in the official build
20:22:44amiconnMikachu: Fixed, about to commit
20:23:03amiconnlinuxstb: Please wear a forehead protector before looking at the diff.. ;)
20:23:34secleinteerok
20:24:03secleinteerexactly what types of patches are you talking about?
20:24:07secleinteerwhat do they do?
20:24:44markunSome added features that are not in the official build. Like album art support for example.
20:25:06 Join Ribs [0] (n=ribs@riblet.plus.com)
20:25:14Febssecleinteer: are you charging via USB? What are you doing while you are charging?
20:25:28 Quit mirak (Read error: 104 (Connection reset by peer))
20:25:28secleinteeri'm charging via wall outlet
20:25:36secleinteerand i'm playing songs while charging
20:25:58secleinteerbattery gets to 100%, and rockbox stops
20:26:02 Join mantono [0] (n=mantono@c83-250-204-173.bredband.comhem.se)
20:26:58secleinteerso i have to do menu+select reboot
20:27:17secleinteerusually i don't have this problem, as i prefer to charge in apple os
20:27:38secleinteerbut in this case, i was also listening to my ipod at the same time
20:27:41markunsecleinteer: I'm sure someone will take a look at it
20:28:59secleinteerok
20:29:01secleinteerthx
20:29:06Mikachuamiconn: i would not have found that error
20:29:50amiconnWell, it only happened on target. Couldn't be an endianess problem, as coldfire is big endian and pp is little endian
20:30:28amiconnAlso couldn't have been a code path thta only works correctly on x86, because I also tested a sim on amd64
20:30:36Mikachumy build system complains apps/plugins/bitmaps/native/SOURCES has no newline at end of file
20:30:38amiconnConclusion: must be iram usage
20:30:48Mikachuclever
20:31:06amiconnAnd putting initialised data (const or not) into .bss isn't a good idea...
20:32:18amiconnaw
20:34:24Mikachuhow come libmpeg2 didn't trigger it?
20:35:12Mikachui can confirm it fixes both my test files
20:36:05markunit fixes the ffmpeg files?
20:36:08Mikachuyeah
20:38:38Mikachubtw, the ffmpeg mpg decoder is a lot faster for me on hdtv videos and such, but it sometimes has bugs
20:39:01Mikachubut maybe it's just sse optimizations and stuff, i have no idea
20:42:55barrywardellamiconn: I updated that OSX patch
20:44:32barrywardellit should affect other builds now
20:44:55barrywardellthe only thing I'm not sure about is this:
20:44:58barrywardell- if [ `uname -m` = "x86_64" ]; then
20:44:59barrywardell+ if [ "`uname -m`" = "x86_64" ]; then
20:45:29barrywardellI mean shouldn't affect other builds
20:48:27linuxstbamiconn: I've just been sleeping, and woke up with the thought "that data shouldn't be in IBSS...". Then I read the logs.... So I would have got there eventually :) But thanks for fixing.
20:48:39 Quit Arathis ()
20:49:19 Join Arathis [0] (n=doerk@p508A6E92.dip.t-dialin.net)
20:53:31dionoeais it possible to sleep less than 1/HZ in rockbox ?
20:53:43dionoea*1/HZ second
20:54:52linuxstbYou have to use the user timer.
20:55:24linuxstbWhy do you need to sleep for < 1/HZ?
20:55:41dionoeai'm trying to playback audio on the ipod's piezo
20:56:32dionoeado you know of any file i could look at for a user time example ?
20:56:49linuxstbI think the lastest piezo driver patch uses it.
20:56:56BigMaclinuxstb: what is the name of the mpeg player the devs are working on
20:56:57dionoeaok, thanks
20:57:13dionoeaBigMac: linuxstb is "the devs" i think :)
20:57:14linuxstbBigMac: "mpegplayer"
20:57:41 Join mirak [0] (n=mirak@ip-7.net-82-216-143.rev.numericable.fr)
21:00
21:00:04 Quit Nibbler ("Leaving")
21:00:49 Quit crash3m|wrk (Read error: 113 (No route to host))
21:01:12BigMacwhats ffmpeg and then?
21:01:20BigMac*and libmpeg2
21:01:56linuxstbffmpeg is a audio and video encoding/decoding library. libmpeg2 is an MPEG-1 and MPEG-2 video decoding library.
21:02:14linuxstbThe Rockbox mpegplayer uses libmpeg2.
21:02:41BigMacok
21:02:47BigMacis there a patch for it yet
21:02:50linuxstbYou can use ffmpeg on your PC to create video files for mpegplayer to play. But at the moment, just the video (no sound).
21:02:53BigMacor is it just starting
21:03:11linuxstbIt's in CVS. Just click on a ".m2v" file in the browser to play it.
21:03:27linuxstb(after downloading the very latest CVS build)
21:03:28 Join Nibbler [0] (n=sven@port-212-202-193-252.dynamic.qsc.de)
21:03:30BigMacok quick question
21:03:43BigMaci have the newest jbuild
21:03:53BigMacif i update my rockbox with the cvs
21:04:10BigMaccan i then put jbuild back over it and play m2v files?
21:04:38linuxstbNo.
21:04:53BigMacwait does this work on the 5g yet?
21:05:00BigMacno right
21:05:00dionoeayes
21:05:05BigMacwoo hoo
21:05:07dionoeait's slow without audio
21:05:25BigMacok linux is the patch in flyspray
21:05:36BigMacbecause ill just apply the jbuild patches along with that
21:05:43linuxstbNo - it's in the Rockbox CVS.
21:05:56BigMacaww
21:06:02BigMacthat meansno video for me:-(
21:06:38markunBigMac: why not?
21:06:54BigMacbecause then i cant have jbuild on
21:07:14BigMacwhich along with the theme im woking on, has all the patches i couldnt live without
21:07:49***Saving seen data "./dancer.seen"
21:07:50linuxstbThat's the problem with using unofficial builds...
21:08:02BigMaci know im getting that now
21:08:04 Join webguest29 [0] (n=550263d8@labb.contactor.se)
21:08:05markunBigMac: why does it matter that it's not a patch then?
21:08:19BigMacbecause i could dl the patches i need
21:08:26BigMacand apply all them plus that one
21:08:31 Join _Lucretia_ [0] (n=munkee@dyn-62-56-49-254.dslaccess.co.uk)
21:08:44linuxstbYou mean apply them to the jbuild source code?
21:09:06BigMacno
21:09:21BigMacohhh
21:09:29BigMacit just clicked in my head
21:09:41BigMacif i patch the source now, itd already have that in it
21:09:48linuxstbYes :)
21:09:59BigMaclol
21:09:59BigMacde de de
21:10:17BigMaclinux your going to bed soon right?
21:10:46linuxstbNot as far as I know...
21:11:08*Febs wishes he could go to bed soon.
21:11:20BigMaclinuxstbamiconn: I've just been sleeping, and woke up with the thought "that data shouldn't be in IBSS...". Then I read the logs.... So I would have got there eventually But thanks for fixing.
21:11:32BigMacthought youd be going back to bed soon
21:11:50BigMacalright then if youll be here im going to try to patch a build
21:14:53 Join Belathor [0] (n=kyle@pool-71-115-141-177.gdrpmi.dsl-w.verizon.net)
21:17:34 Quit secleinteer (Operation timed out)
21:17:45BigMacso linuxst, is video working to a decent extent that you can play through it without it freezing up my ipod
21:17:58BigMacand are the videos going to be very large?
21:18:00 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox 1.5.0.6/2006072814]")
21:18:10Belathorhi! I'm having some trouble installing Rockbox on my 4G iPod via Ubuntu. I am wondering if this is where I could get some help? Thanks!
21:18:32 Join secleinteer [0] (n=secleint@adsl-70-237-232-207.dsl.stlsmo.sbcglobal.net)
21:18:59linuxstbBelathor: Sure.
21:19:10Belathor(=
21:19:10 Join lodesi_ [0] (n=lds@d01v-89-83-171-57.d4.club-internet.fr)
21:19:37 Join leftright [0] (n=leftrigh@p549966DC.dip.t-dialin.net)
21:19:53leftrightFebs ?
21:20:08FebsYes?
21:20:22linuxstbBigMac: The video files are compressed, so they are quite small - I've been encoding at around 450kbps. And yes, video should work reliably, it's just slow at the moment, and there is no sound.
21:20:28BelathorI was following the documentation but I am confused about step 2.
21:20:38linuxstbFirstly, which documentation
21:20:39linuxstb?
21:20:51Belathorhttp://www.rockbox.org/twiki/bin/view/Main/IpodInstallationFromLinux
21:20:59linuxstbOK,.
21:21:24linuxstbWhich part is confusing?
21:21:30BelathorStep 2
21:21:37leftrightFebs: I dont agree with the wordingof Hold Navi in H140 section 2.4.1 WPS key controls, perhaps it might be better to lable it as Press Navi
21:21:43linuxstbBelathor: Step 2 has many small steps...
21:21:54Belathoryeah
21:22:20linuxstbSo which one of those small steps is causing a problem?
21:22:37Belathorkyle@kyle-desktop:~$ ipod_fw -o apple_os.bin -e 0 bootpartition.bin
21:22:37Belathorbash: ipod_fw: command not found
21:22:38leftrightFebs: Hold Navi might be construed as 'select' hold and Navi
21:23:02 Join _birdman_ [0] (n=mrweeeed@cpe-066-057-008-163.nc.res.rr.com)
21:23:12linuxstbBelathor: Did you follow step 1?
21:23:15 Join webguest62 [0] (n=550263d8@labb.contactor.se)
21:23:49Febsleftright: can you give me some context?
21:24:00Belathoryes and there is an ipod_fw executable in my /home/kyle folder
21:24:03Febs(what reference to "Hold Navi" are you talking about?)
21:24:07leftrightFebs: also the picture with descitions of the buttons has the joystick labeled as 'Select', perhpas it would be better to label the joystick as 'Navi"
21:24:30_birdman_hey i need some help
21:24:42leftrightFebs: under getting started, section 2.4.1
21:25:03 Quit _Lucretia_ (Read error: 113 (No route to host))
21:25:10linuxstbBelathor: Then you'll need to type "~/ipod_fw ........." - i.e. give the full path to ipod_fw.
21:25:13 Quit webguest29 ("CGI:IRC (Ping timeout)")
21:25:33_birdman_i just intalled senab's expiramental build, and now i cant get to any menu's all i see is "no files"
21:25:45Febsleftright: OK, I'm with you now. I'll take a look when I have a moment.
21:26:03 Quit Belathor (Remote closed the connection)
21:26:08Febs_birdman_: what is your "show files" option set to?
21:26:10_birdman_cant get to any menu
21:26:33dionoeais the ipod le or be ?
21:26:54Febs_birdman_: and what platform (I'm guessing iPod)
21:27:10_birdman_5gen ipod
21:27:11_birdman_yea
21:27:12leftrightFebs: infact the entire section 2.4.1 there are several references to Hold, does that imply press down and hold or select right and hold without pressing
21:27:57Febs_birdman_: What do you mean "can't get to any menu"? What happens when you press the "Menu" button?
21:28:03mirakis mpeg2 in dailybuild ?
21:28:10 Quit lodesi (Read error: 110 (Connection timed out))
21:28:18_birdman_febs: nothing happens when i press the menu button
21:28:27 Join Belathor [0] (n=kyle@pool-71-115-141-177.gdrpmi.dsl-w.verizon.net)
21:28:38Belathorsorry, got disconnected somehow
21:28:46 Join Davide-NYC [0] (n=chatzill@user-12hdtm5.cable.mindspring.com)
21:29:39Febs_birdman_: is the ipod turned on? if so, what is displayed on the screen right now?
21:29:58_birdman_right now, i just see "no files" and none of the buttons work
21:30:16Davide-NYCHey Febs, if Mmmm shows up (he's on the forums right now) let him know I'm here for testing if necessary
21:30:23Davide-NYC(please)
21:30:42 Quit Arathis ()
21:30:57Belathorlinuxstb: that worked a little better. Now it tells me "Cannot open firmware image file bootpartition.bin"
21:31:03 Join madcow [0] (n=565170b8@labb.contactor.se)
21:31:07Febsleftright: Read section 2.1.1, "Pressing the joystick down is labelled Navi throughout this manual." So yes, you're right, we should correct the photo to be consistent.
21:31:18madcowHi there
21:31:57madcowI'm new here, can someone give me wiki write access. I want to upload a wps to the gallery...
21:32:18 Quit Poka64 ("nite all")
21:32:23miraklinuxstb: I tried the daily build, the video is corrupted
21:32:28linuxstbdionoea: Yes.
21:32:36dionoeathanks
21:32:37mirakthe speed is much better, cpu boost really does a difference
21:32:37linuxstbmirak: Use the latest CVS build.
21:32:53leftrightFebs, still think that Hold Navi is a bit undescriptive
21:32:54Febsleftright: As used in 2.4.1, the word "hold" means a long press of the button.
21:34:17Belathorbootpartition.bin is definitely in my home folder.
21:35:30 Join Arathis [0] (n=doerk@p508A6E92.dip.t-dialin.net)
21:35:31 Join LandusMikain [0] (i=Landus@70-100-181-192.dsl1-erie.roc.ny.frontiernet.net)
21:35:31leftrightFebs: perhaps that should be highlighted or in bold, as to be honest I missed that,
21:35:32Febsleftright: like I said, I'll take a look at it. Perhaps "Long Press Navi" would be more descriptive and less likely to be confused with the hold switch.
21:36:05leftrightLong press is good
21:36:33 Quit Landus (Nick collision from services.)
21:36:39FebsIf we update the picture to be consistent with the text, it will be much more obvious.
21:36:47 Nick LandusMikain is now known as Landus (i=Landus@70-100-181-192.dsl1-erie.roc.ny.frontiernet.net)
21:37:18leftrightPressing the joystick down is labelled Navi
21:37:18leftrightthroughout this manual.
21:38:01BelathorAm I missing something?
21:38:02leftrightFebs;, sorry its that^^ that's what i think should be highlighted
21:38:14madcowIs there an admin in here? (don't wanna bother but need wiki write access).
21:38:28BelathorMy bootpartition is 2GB while the documentation said it would be 40-80 mb
21:38:39 Quit _birdman_ ()
21:38:46Belathor(bootpartition.bin)
21:40:20RoC_MMTha'ts not right.
21:40:44RoC_MMIs your player's storage space 2GB?
21:40:56peturoh madcow... ;)
21:40:59Belathorno, its 20Gbs
21:41:12peturmadcow, what's your wiki name?
21:41:41RoC_MMwhich model player Belathor ?
21:41:54BelathoriPod fourth generation
21:41:59Belathor(grey screen)
21:42:05RoC_MMsame as mine
21:42:56BigMachmm weres paul?
21:43:16RoC_MMMy desktop runs Linux, but I did several installs from various computers running Windows....the 4G has 2 partitions, the first is a 40 MB "firmware/boot/stuff" partition that is never touched...that is what the bootloader instruction are trying to download, patch, and then upload again....the second partition is where all your music and actual rockbox files are.
21:43:17madcowMy wiki name is: jeroen klomp
21:43:27Febsleftright: I understand that you are saying that the sentence should be highlighted, but I don't see how that is necessary when there is a giant picture above it that will contain the correct label for a press of the joystick.
21:43:39madcowThanks in advance
21:44:08FebsAll of the other actions that are possible with the joystick are already labelled.
21:44:12BelathorWhen it asked you put dd if=/dev/sda of=mbr.bin count=1 into terminal, did you put sda or sdb? And if so, what did you substitute for dd if=/dev/sda1 of=bootpartition.bin?
21:44:33RoC_MMWhat is the page for Linux Belathor ?
21:44:43Belathorhttp://www.rockbox.org/twiki/bin/view/Main/IpodInstallationFromLinux
21:44:55peturmadcow: should work now
21:45:05leftrightfebs: on my 12.1 inch screen with pdf zoomed to 100% i have to scroll down and that nice picture is out of view, thats why I suggest that
21:45:05BigMaci can't remember the name of the patching guide paul showed me
21:45:28madcowThanks, we'll see
21:45:51secleinteerok now this is really pissing me off
21:45:55secleinteeri added some songs
21:45:57RoC_MMBelathor, those are two different steps....did the first one...the mbr.bin one make the 2GB file or was that the second command?
21:45:59secleinteerand updated the tag cache database
21:46:07secleinteerand now only 1 or 2 songs out of each album are showing up
21:46:11secleinteerdoes any1 know of this problem?
21:46:13leftrightfebs: plus its such an important operating principal that perhaps it should be highligthed
21:46:28BigMaci think it is that tag cache sucks
21:46:44BelathorRoC_MM: The second one generated the 2GB file
21:46:46secleinteerlol well i really don't have another option
21:46:57BigMacyah you do
21:47:01secleinteerlike?
21:47:08BigMacdo you still have the music?
21:47:12secleinteeryeah
21:47:53Febssecleinteer: what build are you using?
21:47:59secleinteertoday's
21:48:06secleinteerbut this problem has occured with older builds as well
21:48:18RoC_MMBelathor, try to see if cfdisk (partition viewer) is available by doing "cfdisk /dev/sda"
21:48:22secleinteeri think i found a way around it, but i don't remember the exact method
21:48:44BigMacthen make a folder in the root of your ipod and name it music or whatever
21:48:44BigMacjust put all your music in their
21:48:44BigMacso much easier
21:48:59secleinteeri can't do that
21:49:10secleinteerthe filenames for the songs are very inconvenient
21:49:12BelathorRoC_MM: I get a fatal error.
21:49:28secleinteerthey scroll off the screen
21:49:29RoC_MMLike bad command?
21:49:37FebsBigMac: I can think of about a dozen reasons why that is a bad idea.
21:49:43secleinteeri've already tried that method
21:50:08secleinteeris there any way i can debug the problem or something and send a log to the devs?
21:51:25RoC_MMBelathor, you can also view your 4G's partitions by issuing "fdisk /dev/sda"...then when it says "Command (m for help):" enter "p" for "print partition table"
21:51:27secleinteerbtw, where are the irc logs stored?
21:51:32secleinteermaybe i can find my method in them
21:51:34Belathor FATAL ERROR: Cannot open disk drive
21:51:34Belathor Press any key to exit cfdisk
21:51:40RoC_MMhttp://www.rockbox.org/irc/ secleinteer
21:51:44BigMacfebs:such as?
21:51:45secleinteerk thx
21:51:50RoC_MMBelathor, press any key...then follow above instructions
21:51:50 Quit Febs ("CGI:IRC (EOF)")
21:51:59 Join Febs [0] (n=40be240f@labb.contactor.se)
21:52:40madcowCan't edit wiki. maybe wrong password? password reset doesn't work
21:52:41Belathorok
21:52:48Febsleftright: We're probably doing to be adding an overview section to the beginning of the manual that will give a better overall summary of how to navigate Rockbox.
21:52:52BigMacanyone use vmware for compiling?
21:53:23Belathorok, I see my partitions
21:53:28 Quit Daishi (Read error: 110 (Connection timed out))
21:53:34leftrightFebs; ok
21:54:00RoC_MMGreat Belathor ...the "blocks" are probably roughly equivalent to 1K blocks....so if you see a partition of about "40000" blocks that's about 40 Megs and would be the boot partition.
21:54:09excitatorycould someone recommend a tiny, yet clear sans-serif font that works nicely on a nano? i've been using 'snap', which is lovely, but i want to switch it up a little.
21:54:51BelathorRoC_MM: they are all a lot bigger than 40 megs
21:54:57Belathor)=
21:55:04RoC_MMIs the smallest 2 gigs then I suppose?
21:55:07RoC_MMor the first 1?
21:55:09RoC_MMrather...
21:55:20RoC_MM"all"...
21:55:24BigMacexcitatory: are you planning on converting some fonts for rockbox?
21:55:37 Join _Lucretia_ [0] (n=munkee@dyn-62-56-49-254.dslaccess.co.uk)
21:55:40Belathorthe first is 2 gigs and the smallest
21:55:45Belathorits my swap
21:55:56 Join Daishi [0] (n=daishi@ool-45703854.dyn.optonline.net)
21:55:56 Quit webguest62 ("CGI:IRC (EOF)")
21:56:01RoC_MMthe "sda" device must be mapped to another USB device...there are only 2 partitions on a stock 4G partition layout...do you have other USB storage devices plugged in?
21:56:16 Part leftright
21:56:21RoC_MMsda can also map to SATA devices I think.
21:56:42RoC_MMOK
21:57:47excitatoryBigMac: no, but if i knew how, i would consider it for sure.
21:58:20BigMacim sure there is a guide
21:58:24RoC_MMTry this...unplug the iPod from the computer....wait say 5 seconds....then plug it back in, and wait 10 seconds before issuing"dmesg".....alternatively, you could try issueing "dmesg" now, but the hardware detection and device mapping information for the ipod may be a little ways up in the log
21:58:35RoC_MMwhenever you issue dmesg, now or then
21:58:43 Quit madcow ("CGI:IRC")
21:58:51RoC_MMlook for the lines about the ipod, and then look for the part about sda or whatever
21:59:14 Quit merbanan (Read error: 104 (Connection reset by peer))
22:00
22:00:12 Join merbanan [0] (i=banan@farmer.campus.luth.se)
22:00:26 Quit Ribs (".")
22:03:48Belathori think it means that its sdb
22:04:08RoC_MMLikely
22:04:18RoC_MMthat's hte second address it will give out
22:04:22RoC_MMsda is the first
22:04:25BigMacanyone have fontforge on there comp
22:04:38RoC_MMso basically, anywhere on that page (the rockbox page0 it says "sda", just use "sdb"
22:05:18RoC_MMah
22:05:21Belathorok, but do I still use sda1?
22:05:42RoC_MMBelathor, they use fdisk here as well... you can use fdisk -l /dev/sdb to see your ipod partition
22:06:22RoC_MMthe "sd" means SATA or USB, and the "a" means "first device to be detected, "sdb" would be second device to be detected, "sdc" would be 3rd...and the numbers are partition numbers...like
22:06:53RoC_MM"sda1" is your other drive ....and the first partition...and "sdb" is your ipod, the whole thing...and "sdb1" is your ipod's first partition
22:07:04RoC_MMSo no.
22:07:13RoC_MMsda corresponds with your other device...
22:07:30RoC_MMMost of the time there is no "other device" so they didn't point out this caveat I think.
22:07:42RoC_MM(they might have somewhere...I haven't reead the install from linux page entirely)
22:07:52Belathoralright
22:08:23mockerBelathor: Just for some more insight, you can type mount and see where your partitions are mounted, and what they are.
22:08:25Belathorso sdb1 is the 40 meg partition with the firmware on it
22:08:33RoC_MMshould be..
22:10:18RoC_MMSay my name if you have any other questions...my irc will be minimized
22:10:26linuxstbBelathor: You need to be careful when writing the bootloader back - if you don't know which disk device is your ipod, you could accidentally trash your computer's hard disk...
22:11:07BigMaclinux
22:11:25linuxstbBig
22:11:37Mikachumika
22:11:37BigMacdoes linux os have a program thatll convert them to bdf so i can put it on rockbox
22:11:43Belathorthat doesn't sound good. )-:
22:11:58MikachuBigMac: "them"?
22:12:09BelathorI've been trying for a few days without any luck.
22:12:31Belathorok, thanks RoC_MM
22:13:02linuxstbBelathor: It should be straightforward - how many different disks does the command "fdisk -l" (that's the letter "ell") display information for?
22:13:39RoC_MMright
22:14:00secleinteerBelathor: what os are you running? it shouldn't be that hard to figure out which drive is your ipod
22:14:07Spida.oO(the letter e-one-one?)
22:14:11Belathorlinuxstb: sdb1 and sdb2
22:14:14Spida*g*
22:14:23Belathorsecleinteer: Ubuntu
22:14:29secleinteer:D
22:14:41linuxstbThose are two different partitions on one disk - the disk itself is sdb.
22:14:43MikachuBelathor: does ls /dev/disk/by-id give anything?
22:14:47secleinteerthen it should be something like /dev/ipod, i'm guessing
22:15:01secleinteerif it's sdb whatever
22:15:01BigMacfonts
22:15:05secleinteerthen 1 should be your hard drive
22:15:12secleinteeror trying removing your ipod and see what remains
22:15:41linuxstbBelathor: In the "System" column in the output of fdisk, it should list the first one as type "Empty", and the second partition as type "W95 FAT32".
22:16:00MikachuBelathor: you can use the usb-Apple_iPod_00000083065F* devices if you want
22:16:07secleinteerlinuxstb: do you have to use fdisk to get rockbox working with linux?
22:16:12MikachuBelathor: if you press tab, it will fill in the rest of the name for you
22:16:18linuxstbAnd the "Blocks" column for /dev/sdb1 should be something like 40131.
22:16:39linuxstbsecleinteer: I use it to find out which disks have been given which device names.
22:16:52Belathorlinuxstb: yes, thats what I see.
22:16:55secleinteeroh i see
22:17:15secleinteeri think i'll plug my ipod into my ubuntu box just to see what happens :D
22:17:41Mikachulinuxstb: i use ls -l /dev/disk/by-id :)
22:17:48linuxstbBelathor: Then your ipod should be /dev/sdb. So the command to backup the firmware/boot partition is "dd if=/dev/sdb1 of=bootpartition.bin" - that should create a 40MB bootpartition.bin
22:18:11linuxstbMikachu: "fdisk -l" is less typing...
22:18:40Belathorlinuxstb: alright, I'll try that instead of sda1 then! thanks
22:19:07 Join Rudy4Pez [0] (n=A@65-102-20-210.ptld.qwest.net)
22:19:17*mocker | <Belathor has disconnected: Hard drive toast>
22:19:24mocker;)
22:19:30Mikachufitting nick
22:19:40 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox 1.5.0.6/2006072814]")
22:19:42linuxstbMikachu: So I should create symlinks pointing to those? Do you know what the long hex number is after "usb-Apple_iPod" ?
22:19:55Belathorahh! (=
22:19:55Mikachulinuxstb: <tab>
22:21:05Mikachubut yeah, setting up a /dev/ipod* is better for your health in the long run
22:21:31linuxstbI have two ipods though, and I like to mount them both to /ipod (I never connect them both at the same time...)
22:21:34secleinteerMikachu: do you have to edit /etc/fstab to get an ipod recognized in linux?
22:21:41Mikachusecleinteer: uh, no
22:21:47secleinteerk lol
22:21:53Mikachufstab is pretty much for stuff you want mounted automatically at bootup
22:22:03secleinteerthen you do mount?
22:22:12scorcheor if you want to allow a person other than root to mount it
22:22:28Mikachuyou can setup automounting if you fiddle in /etc/hotplug/usb
22:22:28secleinteerscorche: i'd just sudo mount for that
22:22:41scorcheas in a user that is not able to sudo
22:22:42secleinteeri don't think i want to do that :d
22:22:44Mikachuor use something bloated like hal
22:22:45GaloisI noticed that fedora has deprecated hotplug already!
22:22:56amiconnbk
22:22:57dionoea1 or 2 years ago :)
22:23:02secleinteertook me long enough to get my windows share mounted to a local location through /etc/fstab
22:25:04secleinteerwoah
22:25:20secleinteeri plugged in my ipod and rhythmbox pops up right away and recognizes it
22:27:14 Quit excitatory (Read error: 104 (Connection reset by peer))
22:28:55Mikachusee that would annoy me
22:29:03Mikachuif i plugged it in to update mpegplayer :)
22:29:15secleinteermpegplayer?
22:29:17Belathorgreat! now it tells me there is 'no such file or directory' as sdb1
22:29:36secleinteerBelathor: it shows up for me as /media/ipod
22:30:47dionoeaBelathor: could you paste the contents of fdisk -l /dev/sda and fdisk -l /dev/sdb on pastebin.com ?
22:30:48Belathor): I don't see it there
22:30:59secleinteerwhat do you see under /media?
22:31:14Mikachusecleinteer: we're not even trying to mount it...
22:31:56secleinteeraren't you trying to find out which one it is?
22:32:59secleinteerwhether it's sda1 or 2?
22:33:17Mikachuhah, that was never the question
22:33:26secleinteeroh lol ;0
22:33:47BHSPitMonkeysecleinteer, try to avoid giving Ubuntu-specific advise, too... :P
22:34:10Belathorsecleinteer: cdrom, cdrom0, sda2,sda3, usbdisk
22:34:29secleinteerBHSPitMonkey: well, the only reason i'm doing that is because belator is running ubuntu
22:34:39secleinteerBelathor: maybe usbdisk?
22:34:55Belathorsecleinteer: fdisk -l /dev/sdb is taking forever
22:35:28secleinteerok.....is that good or something?
22:35:35secleinteercan't u cancel it?
22:35:56BelathorI don't think its good.
22:36:04BHSPitMonkeyCTRL+C
22:36:07secleinteerok
22:38:35dionoeaamiconn: how did you find that video corruption fix ? (seems like black magic to me :D )
22:38:46Mikachuhe explained it to me earlier
22:39:02*dionoea browses up
22:39:09amiconndionoea: Check the log around 20:29
22:40:16 Join tommeyer [0] (n=icechat5@cpe-24-161-131-252.hawaii.res.rr.com)
22:40:22dionoeaindeed ... (the 3 arrays look like const data to me though)
22:40:23 Part tommeyer
22:40:25 Join tommeyer [0] (n=icechat5@cpe-24-161-131-252.hawaii.res.rr.com)
22:40:51amiconndionoea: The 2nd and 3rd have no 'const' on them, hence they need to go into .idata
22:41:04amiconnOtherwise compiling produces a section type conflict
22:41:12dionoeahum ... ok :)
22:41:37*linuxstb likes to do these things to test amiconn's debugging skills...
22:41:44dionoea:D
22:41:58*amiconn is searching for methods to speed things up...
22:42:18Belathorsecleinteer: It was found by the system automatically and appeared on my desktop. now it doesn't do that anymore
22:42:50linuxstbBelathor: Have you tried unattaching your ipod, and then reattaching it? It sounds like your computer has lost contact.
22:43:08tommeyerHi, I've been trying out mpegplayer on the video ipod, and it's not all that fast. I'm happy to help speed it up, if anyone can point me in the right direction.
22:43:24linuxstbtommeyer: You'll need to find your own direction...
22:43:52Belathorlinuxstb: yeah, but its not being detected by the os as an iPod anymore
22:43:53amiconnlinuxstb: I presume you didn't import all of libmpeg2. Can you point me to viewcvs or sth?
22:43:53BigMachmm i have a temporary fix
22:43:54dionoealinuxstb: did you commit the ipod video yuv blit function yet ?
22:44:55linuxstbamiconn: dionoea pointed me here earlier: svn://svn.videolan.org/libmpeg2/trunk
22:45:12linuxstbBut I downloaded it from http://libmpeg2.sourceforge.net
22:45:18linuxstbdionoea: Not yet, no.
22:45:24*amiconn has no svn installed
22:46:05linuxstbtommeyer: There's one optimisation I haven't committed to CVS yet for the 5g - as dionoea mentioned (the yuv blit function).
22:46:23BigMaclinux
22:46:36BigMacok so what do i need to watch my videos on my 5g
22:46:45BelathorI'm going to log out and log back in.
22:46:48 Quit Belathor ("Leaving")
22:46:49BigMacthe converter and what else
22:46:59amiconnsourceforge is acting up :(
22:47:21dionoealibmpeg2 isn't hosted on sourceforge anymore (at least not the repository if that's what you're looking for)
22:47:37BigMacamiconn: what programs do i need to get my video into a rockbox watchable form
22:47:39dionoeai'll add a libmpeg2 trac so that people can browse online
22:47:41dionoea2 secs
22:48:19 Join qwx [0] (n=qwm@h229n1fls34o1010.telia.com)
22:48:55BigMacdionoea: how do i convert to m2v?
22:49:29 Quit qwm (Read error: 110 (Connection timed out))
22:49:30dionoeai can tell you using VLC
22:49:40dionoeabut other people here us mencoder
22:50:03BigMacwhich one would be easier to use
22:50:07BigMacuser friendly
22:50:26 Join LandusMikain [0] (i=Landus@70-100-181-192.dsl1-erie.roc.ny.frontiernet.net)
22:51:51amiconnlinuxstb: I know why the framerate display is a little high first, and then goes down. (The effect is more obvious with low framerates)
22:52:11amiconnYou're not counting the time to decode the very first frame, but you count that first frame
22:52:59BigMacdionoea: im trying to go from either mp4 and mov to m2v
22:54:49linuxstbamiconn: Ah, I knew it would be something stupid...
22:55:25amiconnI'll change that, and make the fps display fps independent
22:55:48amiconn(currently it displays every 125 frames, I'll make it display every 2 seconds)
22:56:44amiconnmeh
22:56:52*amiconn doesn't like code w/o whitespace
22:56:56dionoeausing VLC it's: vlc <input> −−sout "#transcode{vcodec=mp2v,vb=600,width=<the width>,height=<the height>}:es{access-video=file,mux-video=es,dst-video=<filename>.m2v}"
22:58:51BigMacdionoea: with quotes?
22:59:11dionoeaunder linux yes
22:59:21dionoeaelse the {, }get expanded
22:59:32BigMacdoes it work in windows?
22:59:46dionoeasure
23:00
23:00:00BigMacand width is 240 on the video ipo and height is 320 correct?
23:00:06dionoeajust remove the quotes and add an = sign between sout and #transcode
23:00:17BigMacok
23:00:24dionoeaor do it in the interface and cut and paste the sout chain
23:00:34BigMacand once i install vlc, i type this directly into the cmd?
23:01:05dionoeacd "C:\Program Files\VideoLAN\VLC"
23:01:09dionoeaand then type this command
23:01:15dionoeaor use the interface
23:01:23 Join Belathor [0] (n=kyle@pool-71-115-141-177.gdrpmi.dsl-w.verizon.net)
23:02:56dionoeaamiconn: there you go http://trac.videolan.org/libmpeg2/browser
23:03:36amiconnlinuxstb: The rate still goes down...
23:03:58BigMacdionnea:does vlc support mp4 or .mov
23:04:03dionoeasure
23:04:18BigMacyes?
23:04:20dionoeaVLC supports everything except real media formats
23:04:27BigMack
23:05:41BigMacgod i hate my internet
23:05:51BigMacit wont load crap anymore
23:06:08Bagdernot loading crap sounds like a good idea to me
23:06:12Bagder:-)
23:06:20dionoealol
23:06:46BigMac:-0 i like my crape
23:07:25BigMacdionoea: how do i know when its done
23:07:34BigMaccmd popped up then dissapeared
23:07:43BigMacbut its not in program files
23:07:49RoC_MMI think you mean, Real Media can't be supported by anything because it's a secretive spec.
23:07:51***Saving seen data "./dancer.seen"
23:07:53dionoeayou're in the installation directory ? (/me hates windows command line)
23:07:53 Quit lodesi_ ("leaving")
23:08:15BigMacyah
23:08:22dionoeawhat command did you use ?
23:08:27BigMacnone
23:08:31BigMacthe exe opened
23:08:34 Quit Landus (Nick collision from services.)
23:08:35dionoeahehe
23:08:36 Nick LandusMikain is now known as Landus (i=Landus@70-100-181-192.dsl1-erie.roc.ny.frontiernet.net)
23:08:37BigMaccmd flashed up
23:08:41BigMacthen dissapeared
23:08:54dionoeathe cmd.exe exe ?
23:09:03BigMacno
23:09:19BigMachttp://www.videolan.org/mirror.php?mirror=http://mirrors.optralan.com/videolan/&file=vlc/0.8.5/win32/vlc-0.8.5-win32.exe
23:09:27merbananRoC_MM: well lots of it has been reverse engineered
23:09:39dionoeaoh ... the installer :)
23:09:48RoC_MMReal does not suppor reverse engineering.
23:09:59RoC_MMor rather
23:10:04dionoeacook audio was reverse engineered
23:10:05RoC_MMReal does not support compatibility
23:10:06dionoeabut not video
23:10:19RoC_MMI like that best
23:10:26RoC_MMReal does not support compatability
23:10:42dionoeaBigMac: the installer doesn't run ?
23:10:52BigMaclike i said
23:10:58 Quit Bger ("nite")
23:11:01BigMaca command prompt opened
23:11:01BigMacthen closed
23:11:07BigMacthen nothing else came up
23:11:20linuxstbamiconn: Looking at your H300 lcd_yuv_blit, was there a reason you check that width and height are even, rather than forcing the app to do it?
23:11:33 Join bluebrother [0] (i=4oyoLl32@nat-wh-1.rz.uni-karlsruhe.de)
23:11:44 Quit Spida (Read error: 131 (Connection reset by peer))
23:11:53dionoeayou're like the first person in 10 milion to have a problem with that installer :) (or i'm not understanding what you're telling me)
23:12:17BigMacwell this sucks
23:12:33dionoeayou're sure that you have the full file ?
23:12:35amiconnlinuxstb: I don't check it, I force it to be even
23:12:41linuxstbYes, that's what I meant.
23:12:56BigMaci dont have the program installed yet
23:12:57amiconnIt's just to not confuse the lcd controller (for width)
23:13:02 Join Spida [0] (n=timo@213.164.155.67)
23:13:05amiconnFor height it's in fact unnecessary
23:13:10BigMacso i dont think ill need the line
23:13:38dionoeayou could just download the zip and unzip it anywhere if the installer is a problem
23:13:47dionoeabut double clicking on the installer usually launches nsis
23:15:29RoC_MMdownload it agan
23:15:52 Quit Hansmaulwurf ("( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )")
23:15:59dionoeaand don't use the optralan mirror, it has crappy upload
23:16:52mirakths cvs build are automatically done by the server ?
23:18:21Bagdermirak: cvs builds are built by a whole cluster of servers
23:18:23dionoeaBigMac: is it working now ?
23:21:03mirakdoes anyone volonteered for assembly optimisations of idct yet ?
23:21:31BigMacwhich do you suggest?
23:21:34mirakI will go camping for 1 week
23:21:43dionoeaBigMac: the first one in the list
23:22:10BigMacunder exe?
23:22:15BigMacor 7zip
23:22:17dionoeamirak: you code assembly when camping ?
23:22:29dionoeaBigMac: exe if you want to run an installer, 7zip or zip if you don't
23:22:36dionoeayou choose
23:22:51BigMaci code assembly while im in class
23:22:55mirakdionoea: on the beach on laptop by 40 degrees at shadow of the sun
23:22:59dionoea:)
23:24:38miraksince idct is n² and bidimentional we would probably increase performances much significantly than for audio codecs
23:24:55mirakno ?
23:24:59 Join S0ap [0] (n=bob@unaffiliated/s0ap)
23:25:00dionoeayeah! 23 fps on ipod video with 100x130 resolution
23:25:23dionoeai should've selected a multiple of 8 size
23:25:29dionoeaborders look weird
23:25:50amiconnYou want multiples of 16
23:25:55*linuxstb commits the lcd_yuv_blit for the ipod video.
23:26:11*dionoea runs CVS up
23:26:28*linuxstb forgot to run CVS up
23:26:44dionoea:(
23:27:31BigMacwhats a good windows program that yall suggest
23:27:31BigMac\
23:27:46S0apfor...?
23:28:09BigMacim not sure
23:28:14BigMacanything
23:28:17linuxstbdionoea: OK, it should be there now.
23:28:19BigMacsome good programs
23:28:21BigMacuseful
23:28:25dionoealinuxstb: neat :)
23:28:25Bagdercurl!
23:28:26BigMacnot frree
23:28:26Bagder;-)
23:28:30bluebrothergvim
23:28:34linuxstbipodpatcher.exe
23:28:34BigMaccrul?
23:28:37dionoeaformat
23:28:43linuxstbrockboxui.exe
23:28:55bluebrotherformat.com ;-)
23:28:59linuxstbvmware.exe
23:29:10BigMaci have vmware
23:29:13bluebrotherloadlin.bat
23:29:52BigMacwow you guys are funnnnnnny;-)
23:29:59markunBigMac: foobar2000
23:30:01 Join stripwax [0] (n=stripwax@i-83-67-214-206.freedom2surf.net)
23:30:05BigMacnot free
23:30:07stripwaxyo
23:30:15BigMaci already have it
23:30:18S0approcess explorer / rootkit revealer / AVG antivirus / AutoCAD LDD...
23:30:35BigMacavg is free
23:30:41dionoeaMS excel
23:30:46BigMacand what do the others do
23:30:52peturirfanview
23:31:05BigMacand btw dont bother suggesting any macromedia programs i have them all
23:31:05peturpicasa
23:31:08*stripwax wonders what was the question?
23:31:17S0apdang you must be rich BigMac.
23:31:21BigMacpetur:ive said not free like 12 times
23:31:47BagderBigMac: you may notice that not all suggestions were 100% serious
23:31:58BigMacstripwax: good non free programs
23:31:59Bagderso what you said has little importance ;-)
23:32:01BigMaci got that
23:32:06dionoeaformat isn't free ...
23:32:15BigMacs0ap:no just kind friends
23:32:20stripwaxOS/2 Warp
23:32:25scorcheBagder: is it ever anything else? ;)
23:32:32Bagderno
23:32:37dionoeaBigMac: you also need Nero
23:32:40Bagderbut sometimes you need to clarify
23:32:43BigMacbrb
23:33:51stripwaxI use very little non-free software, but I'd name Paint Shop Pro, Partition Magic, Nero and NTrack Studio as things I have paid for.
23:34:18dionoealinuxstb: 26 fps instead of 23 fps for my 100x130 sample. That's quite an improvement :)
23:34:24amiconnlinuxstb: What about optimisations that don't change speed, but save IRAM?
23:34:52dionoeastripwax: you paid for Partition Magic and Nero ?!?
23:35:02linuxstbamiconn: What do you mean? Removing some things I've already put in IRAM?
23:35:03amiconn(the CLIP() thing in idct.c)
23:35:11stripwaxdionoea - yes, because they are non-free.
23:35:34amiconnNo I mean implementing the clipping as a tiny asm inline instead that 7.5KB table lookup
23:35:59amiconnAlmost the same thing as in jpeg.c - the clipping is just 3 insructions on coldfire
23:36:19amiconnI tried it, it doesn't change speed at all
23:36:27linuxstbSounds good to me. Can you do it for ARM as well?
23:36:54amiconnI can, but I can't test the speed impact
23:37:11linuxstbI'm happy to test.
23:37:27linuxstbOr you could run it on your mini, disabling video output. I've done that on my H140.
23:39:52 Join Variable [0] (n=A@65-102-20-210.ptld.qwest.net)
23:40:11 Join excitatory [0] (n=excitato@CPE-70-94-13-227.wi.res.rr.com)
23:40:17tommeyerlinuxstb: Just tested your 5g patch, get 26fps on a 176x132 video, and 8fps on a 320x240
23:41:30amiconnlinuxstb: Then I can't see whether it is working correctly
23:41:31FebsDamn it linuxstb. I just updated my ipod video build an hour ago. Now I have to do it again!
23:41:43FebsYou guys have been FAR too productive recently.
23:42:10Febs:)
23:46:16mirakamiconn: you can extract images to drive
23:46:32 Quit BigMac (Read error: 60 (Operation timed out))
23:46:45mirakwell, there is what is needed to do that in mpeg2dec
23:47:05 Quit Rudy4Pez (Read error: 110 (Connection timed out))
23:49:10linuxstbtommeyer: Do you know ARM assembler?
23:49:18 Join rotator [0] (n=e@rockbox/developer/rotator)
23:49:46linuxstbFebs: I wouldn't call a video player being productive :)
23:51:30FebsSeems the masses would disagree with you.
23:52:40S0apthe masses need to wash
23:53:14stripwaxhrmmm.. is anyone closer to figuring out how the broadcom chip or dma works on ipod 5g?
23:53:24*stripwax wonders if he needs to wash..
23:55:16linuxstbstripwax: AFAIK, no-one in Rockbox is doing any ipod reverse-engineering. Although Bagder's work on the Sansa may help.
23:55:43stripwaxlinuxstb - ah yeah, good point.
23:56:03linuxstbIf you look at the IpodStatus page, there are still a lot of things to do that we know how to do....
23:57:06stripwaxso, this video player - does it have known bugs or issues other than speed?
23:57:48linuxstbI don't think so. Ask the masses...
23:58:06tommeyerHow close is support for the second processor? That would reduce need for optimizations if it's close
23:58:55linuxstbThe COP won't reduce the need for optimisations - we need that for audio decoding and disk buffering...

Previous day | Next day